Sage Journals: Discover world-class research

Abstract

Severe safety accidents occur frequently in underground mines. It is crucial to develop safe mining technology. Smart mining is a reliable solution to this issue. The unmanned electric locomotive is one of the key components in smart mining. Literature review indicates that there are only six mines around the world that have equipped with unmanned electric locomotives, and very limited works have discussed the challenges in this field. Hence, this work aims to discuss the challenges. The first concern was underground position tracking. A solution based on wireless sensor network was introduced. Three challenging issues in wireless sensor network, that is, sensor node distribution, energy supply, and efficient data transmission, were discussed. The second concern was the control system of unmanned electric locomotives. The challenging tasks in this concern were discussed. First, the nonlinear dynamic response of the interaction between the rail and locomotive may influence the control performance of the unmanned electric locomotive. A coupled mathematical model describing the interaction may help the controller design. Second, the speed sensorless control strategy was introduced. The possible future research directions were then summarized. The discussions in this work will provide reference for optimum control system design for unmanned electric locomotives in underground mines.

Keywords

Smart mining unmanned electric locomotive underground position system wireless sensor network sensorless control

Introduction

Mineral energy is a pillar of the national economy in China. Currently, the coal energy accounts for 73% of the national energy production and consumption, and in the next 50 years, it will still account for more than 50% of China’s energy consumption.¹ According to the “China Resource Efficiency: Economics and Outlook” issued by the United Nations Environment Programme (UNEP) in 2013,² China’s consumption of raw material, including the construction mineral, metal mineral, fossil fuel, and biomass, is the number one in the global market. In 2010, China consumed about 20% of the world’s energy, 60% of cement, 47% of iron ore, 49% of steel, 44% of lead, 40% of aluminum, and 38% of copper. In 2012, China invested 1259.24 billion RMB in geological and mineral exploration and 1.31 trillion RMB in fixed assets of mining industry.³ About 227 new mineral resources were discovered, including 109 large mineral areas.³ These statistics indicate that in the 21st century, mineral resources of China are still in the leading position, and the mining industry is still the basis for the sustainable development of the national economy. Meanwhile, China’s mine production continued to maintain steady growth.² Consequently, the mining industry will remain as the foundation of national economy in China.¹

However, the development of advanced mining technology is slow. Severe accidents frequently occur in underground mining areas. According to the statistics presented by China Administration of Work Safety,⁴ in 2004, there are 3639 mining catastrophes in China and 6027 people lost their lives; the numbers are 3341 and 5986 in 2005, respectively. Up to date, there are at least three miners died each week in China.⁵ These catastrophes have brought considerably negative influence to the sustainable development of the mining industry.⁶ Hence, it is imperative to develop advanced mining technology to reduce mine accidents.

Smart mining technology is an effective way to solve the safety problem in mines.⁷ Smart mining is one of the most important research fields in the development of the next-generation mining technology. By means of digitalization and informatization, smart mining takes advantages of the active sensing, remote signal/data communication, and unmanned/automatic control to reduce the human factors in operating mining machines and improve the efficiency and safety of the mining process. However, some key technologies related to the smart mining are still under development and evaluation. One of them is the unmanned electric locomotive. Since 70% of mining accidents occurred in the mining area, the unmanned electric locomotive technique can reduce the human-injury accidents and hence facilitate the sustainable and healthy development of mining technology.⁸ To date, the unmanned electric locomotive systems have been tested in only six mines over the world.⁹ The control performance of the unmanned electric locomotive would be significantly influenced by the interaction between the operating environment and locomotive. The most influential factors for the unmanned electric locomotives include adaptive–accurate position tracking, efficient signal/data communication, and high-performance remote control techniques. However, very limited work has reported these issues. To this end, this work focuses on these key factors that determine the control performance of unmanned electric locomotives in underground mines. Two challenging tasks are discussed from the aspects of (1) construction of underground position system for locomotive position tracking and (2) advanced controller design for unmanned locomotives. Some potential solutions and suggestions were summarized.

The remainder of this article is organized as follows. Section “Progress on unmanned electric locomotive” reviews the progress on unmanned electric locomotives in smart mining. The challenges in accurate position tracking using underground position system are introduced in section “Challenges on position tracking in underground mines.” Section “Challenges in reliable controller design” describes the challenges in high control performance of unmanned electric locomotives. Then, section “Future research directions” discusses the future research directions for unmanned electric locomotives. Finally, section “Conclusion” presents the conclusions of this work.

Progress on unmanned electric locomotive

Unmanned electric locomotive system is one of the key systems in the smart mining. An unmanned electric locomotive system mainly consists of four parts. These four parts include the variable frequency control system, the wireless communication system, the automatic scheduling, monitoring and protecting system, and the power supply and control system. Different from ground automated load–haul–dump (LHD) vehicles, the unmanned electric locomotives are designed for underground mines (e.g. coal mines and/or metal mines). Hence, because of hashing and unknown working environment in the underground areas, the sensing and control system of the unmanned electric locomotives is much more complicated than that in the ground LHD vehicles.

At present, most of the electric locomotives used in the underground rail transportation are controlled by drivers.¹⁰ Due to harsh environment and human factors, the mine dispatching center cannot communicate timely along with the drivers, leading to low scheduling efficiency. Moreover, accidents of the vehicles (such as collision) occur frequently. With the continuous development of mining technology and the demand expansion of mineral resources, there are more and more large mines with over 10 million tons of annual production.⁷ Reliability, efficiency, and safety are the common requirements for the underground transportations.¹¹ The unmanned electric locomotive system is a promising solution to these requirements. It can not only enhance the automation degree of ore transportation, but also prevent severe accidents. No matter for large mines, super large mines, or special mines, the unmanned electric locomotive can significantly improve the safety, economic and environmental benefits, and social benefits of mineral production. Consequently, the unmanned electric locomotive has been regarded as a new-generation technology.¹²

Russia developed a prototype of unmanned electric locomotive in 1969.¹³ The radio signal was used to realize the remote control. The Kiruna mine in Sweden implemented unmanned railway transport plan in the early 1970s and put unmanned electric locomotive into use in 1980.¹⁴ Then, in the late 1990s, the remote control technique was developed for the mining machines in the Kiruna mine,¹⁵ and the unmanned walking system¹⁶ and automatic mining system¹⁷ were applied in the Bell Allard mine and Brunswick mine applied in Canada. In the 1990s, Australia’s Commonwealth Scientific and Industrial Research Organisation (CSIRO)¹⁸ began the development of smart mining and devoted to advanced sensing system and remote control system in the mining and loading operations. Some of their technologies have been put into commercial products by Caterpillar.¹⁸ On the same period, the University of Helsinki in Finland started to develop the smart mining technology. A 5-year’s plan was conducted to modernize the Finnish mining group (i.e. Outokumpu Oy, Tamrock Oy, Normet Oy, and Lokomo Oy) in terms of mining machinery modernization, real-time monitoring communication, computer-aided information management, and automatic control.^12,19 Another 5-year’s plan was also scheduled to realize the remote monitoring and control for the mining machineries. Now, early harvests of this plan have been obtained in the unmanned mining in the Kemi mine of Outokumpu Oy. In these plans, the focus was given to the design of fully automated mining machines.¹² The International Nickel Corporation (INCO) spent 12 years to practice the automatic mining technology and improve its productivity by 250%.²⁰ In the light of this output, the Rio Tinto Group announced in 2012 that they would invest 5.18 billion dollars in building an unmanned electric locomotive system in order to improve the productivity of the Pilbara iron ore.²¹ The first unmanned electric locomotive was planned to put into practice in 2014, and the whole project was expected to complete in 2015. The goal of this project aimed to improve the capacity of railway network, save energy, and reduce emissions. In addition, the Rio Tinto Group would invest in developing unmanned trucks and drilling machines to further support the unmanned mining system.²¹ In 2011, China ENFI cooperated with Dongguashan Copper Mine of Tongling Nonferrous Metals Group to design the first unmanned electric locomotive in China.²² The capacity of this locomotive was 20 t. The computer control technology, bus technology, and wireless communication technology were comprehensively used. Moreover, the intelligent control was employed. In 2013, this unmanned electric locomotive system was officially put into practice in the Dongguashan copper mine. It was expected that the energy consumption could be saved above 30%.²² Besides China, many countries are also preparing for the unmanned electric locomotive system. In South Africa, the automatic mining technique has been introduced into many mines to improve the mining safety and efficacy.²³ The remote control technique was adopted by the Mt Isa in Australia, Saskatchewan uranium mine in Canada, and El Teniente copper mine in Chile.²⁴ With the help of remote sensing technology, the mining efficiency was improved greatly in these mines. Hence, they planned to introduce the unmanned electric locomotive system in the near future.

Although the unmanned locomotive can significantly improve the efficiency and safety in the underground mining transportation, the existing research results indicate that it still needs to overcome some key technologies for the unmanned electric locomotive to promote its development and application in underground mines. Since in the underground, the locomotive would frequently turn on and off, it requires high control precision and stability for the control system. High-performance controllers are therefore the key technology to ensure the smooth operation of unmanned electric locomotives. The challenging issues in performing high control performance for unmanned electric locomotives are discussed in the following sections.

Challenges on position tracking in underground mines

1. Position tracking is an enabling capability for operation of autonomous vehicles.

In order to control the electric locomotives in an unmanned manner, the first issue is to precisely track the position of the locomotives. This is because positioning information is used by schedulers, for motion control and for collision avoidance (interaction with other vehicles and infrastructure). Implementation may take various forms, such as positioning relative to local features or positioning within a global map. The local positioning system is usually costly, while it is much cheaper and easy to construct the global map–based positioning system. Global positioning system (GPS) can be a promising choice.

2. Surface applications have leveraged satellite-based GPS. Augmented systems in use where satellite visibility is reduced or quality of service requirements dictate particular accuracy and availability.

As well known, GPS, supported by a series of orbiting satellites, has been widely used to provide powerful navigation service for ground vehicles.²⁵ The movement trajectory of a ground target can be effectively and precisely tracked by the satellite-based GPS. The location, speed, and direction of the object can be obtained. However, it is always inefficient and inaccurate for GPS to track the positions of underground vehicles because the line of sight is necessary for the satellites.²⁶ The performance of GPS is rather poor for locating underground vehicles since GPS signals cannot penetrate into the ground.²⁷

3. Underground positioning may use terrestrial beacons to extend the signals of conventional GPS but still in early days.

In order to address the limitation of GPS in underground applications, the underground positioning system has attracted considerable attention in the communications engineering.²⁸ Generally, underground positioning adopts extremely low-frequency electromagnetic wave to carry the GPS signals. Low-frequency waves can travel a long distance underground. To receive the GPS information, the low-frequency electromagnetic wave can be launched either from ground GPS equipment to underground GPS receiving end or from underground to ground. Swiss Institute of Speleology and Karstology (ISSKA) has started to develop its underground positioning system since 2004.²⁹ The developed prototype Underground-GPS (U-GPS v2) has been tested and is able to follow a caver in real time. The DigiTrak Eclipse³⁰ is another useful underground locating system, but it is initially designed for horizontal directional drilling (HDD).³¹ For underground mining, several attempts have been made for the purpose of effective locating of mining equipment. Chen et al.³² used the radio frequency (RF) identification devices (RFID) for underground positioning. Maqsud et al.³³ adopted the micro-electromechanical systems (MEMS) to construct the underground mine positioning system. Conceicao et al.³⁴ employed the wireless sensor network (WSN) to establish underground GPS. Liu and He³⁵ used the chirp spread spectrum (CSS) to improve the position accuracy of WSN-based underground positioning system. Salih et al.³⁶ suggested ultra wideband (UWB) for underground positioning. Liu and Wang³⁷ combined WiFi and UWB to provide real-time underground position. Cypriani et al.²⁸ tested WiFi-based position system for a 70-m underground gold mine. Katayama³⁸ presented the GPS repeater system for underground mall. Chen and Xiao³⁹ proposed the ZigBee-based underground positioning technique. Baffour and Abatan⁴⁰ discussed the possibility of combination of ground penetrating radar (GPR), geographic information system (GIS), and GPS for underground position system. Among these existing publications, WSN is the most promising technique for underground positioning system.⁴¹ However, there are still several unsolved challenging issues before developing a reliable WSN-based underground positioning system.⁴²

4. Challenges in constructing underground positioning system using WSN.

(a) Data collision problem: the underground positioning system requires enough sensor nodes to cover the entire operating mining area. The increase in sensor number will result in decrease in WSN reliability. Hence, it should optimize the installation positions of the sensor nodes to control the sensor number. GIS may assist the sensor-installation optimization by providing useful underground geographic information.³⁹ However, how to incorporate the GIS information into the configuration of WSN is a challenge.

(b) Power supply problem: WSN is powered by batteries; however, it is always difficult to replace new batteries in underground mining area. As a result, the battery issue needs to be solved in underground positioning system. The energy harvesting is regarded as a useful technology to address this issue. However, the popular energy-harvesting techniques, such as solar-, piezo-, and thermal-based methods, are subject to a common limitation, being reliant on ambient sources that cannot be accessed in underground area. The recent RF energy-harvesting technique^43,44 can be used to recharge wireless sensors in inaccessible or hazardous areas and is able to provide maintenance-free power for a large amount of sensor nodes. The current research hot topic is to find a suitable way to construct RF energy harvesting into supercapacitor, thin film, or solid-state batteries. Another challenge is to develop suitable RF-based recharge process for lithium-ion battery to avoid unexpected damage/aging because of unreasonable recharging.

(c) Efficient data transmission: the data transmission rate is very low in underground areas. If image and/or video data are required, the transmission efficiency and energy consumption will be unaccepted in underground position system. Fortunately, compressed sensing (CS)⁴⁵ is a promising technique to solve this problem. CS overcomes the bandwidth limitation in Shannon–Nyquist sampling theorem by sparse sampling, which greatly reduces the signal sampling rate and data storage. Currently, CS has been applied in various applications,⁴⁶ including nuclear magnetic resonance, radar signal, image processing, and WSN. However, CS requires sparsity/sparse representation of the analysis signal.⁴⁵ Some sparse transformation base (i.e. Fourier, wavelet, curvelet, and contourlet transformation bases) can be used to sparsify a simulated signal,⁴⁷ but their performance is poor when sparsifying practical signals. How to deal with a practical signal that is not sparse or cannot be sparsified is a challenge. Pioneer work has been found in variation constraint,⁴⁸ block sparsity,⁴⁹ Wavelet-tree,⁵⁰ markov-tree,⁵¹ local self-similarity,⁵² and sparse Bayesian learning.⁵³ These methods rely on strong mathematical hypothesises that are not applicable in practice. A de-noised approximate message passing (D-AMP)⁵⁴ makes it realistic to sample a practical signal using CS. However, very limited work has been done to address D-AMP in CS, and the application of D-AMP-based WSN has not yet been found.

Therefore, if the aforementioned challenging issues can be addressed, an efficient and reliable WSN-based underground position system can be expected for the unmanned electric locomotives.

Challenges in reliable controller design

The vector control and direct torque control are often employed in the variable frequency inverters. The vector control technique achieves the torque control of the driving motor by decoupling control of the motor stator current. Its control dynamic quality is good, but it consumes a large amount of calculation. The direct torque control technique directly controls the torque of the driving motor. Though its control computation is less than the vector control, it often subjects to the torque ripple in high and low speed.²² Therefore, the control performance of the vector control is better than the direct torque control in applications where the speed/position adjustment accuracy is strictly requested. Given that the underground electric locomotive is requested to stop at the designated locations or to run at reference speeds, it is reasonable to adopt the vector control for the unmanned electric locomotive systems. It should be noted that the controller design using the vector control technique is always closely associated with the working environment of the control objective. There is little research that addressed the control of underground unmanned locomotives. Hence, in order to obtain high control performance, it is crucial to construct a robust controller based on vector control for the unmanned electric locomotive system according to the specific working environment. The major challenges in reliable controller design for the unmanned electric locomotive system are discussed in the following.

1. The first challenge involves the mathematical model describing the interaction between the underground rail and locomotive body motion.

The complicated topography and rapidly changeable environment cause strong vibration on the electric locomotive body, leading to speed fluctuation of the locomotive. Moreover, the locomotive operates with frequent start–stop and acceleration–deceleration commands. The large inertia of the locomotive body directly affects the operation of the driving motor. That is to say, the control system should consider not only the speed control of the driving motor but also the thrust of the locomotive body. Thus, the coupled movement of the locomotive and the driving motor constitute as an organic whole. On one hand, the vibration and inertia of the locomotive body affect the reasonable allocation of the driving motor speed and torque. On the other hand, the speed and torque control of driving motor directly determine the running speed and vibration of the locomotive body. Hence, the dual coupling effect between the locomotive body and the driving motor raises the challenge on multi-body decoupling control of both the driving motor and the locomotive body. This control problem belongs to the multi-input multi-output (MIMO) control, which is a hot topic in the field of vehicle control. However, little research has reported the MIMO control of the underground electric locomotive. Therefore, from both academic and practical aspects, it is imperative to investigate the MIMO controller for the decoupling control of the motion of the locomotive body and driving motor.

In order to design suitable MIMO controller, it is necessary to establish the mathematical model about the interaction between the underground rail and locomotive body. This is because the vector control technique requires effective identification of the control model parameters, including resistance, inductance and flux of the driving motor, and the speed and vibration of the locomotive body. But due to strong nonlinearity of the unmanned electric locomotive system, it is difficult to establish precise mathematical model of the underground rail-locomotive body-driving motor. Consequently, it is inefficient for the traditional controllers to obtain high control performance. For instance, the traditional proportional–integral–derivative (PID) controller is difficult to meet the control requirement of time-varying nonlinear systems.⁵⁵ Although some industrial control methods, including the sliding mode variable structure control, robust control, and back stepping control,^56,57 can deal with nonlinear control systems, they still need to overcome the model boundary problem of the control system. However, the model boundary is determined by a comprehensive understanding of the mathematical model of the control system. Several intelligent control methods can solve the model boundary problem. These methods include the fuzzy control and neural network control.⁵⁷ They can achieve excellent control performance without knowing the exact mathematical model of the control system.

The aforementioned industrial and intelligent control methods are applicable to their own unique control conditions and scenarios. Neither of them can be used for all control objects. The latest study^58,59 strongly suggested that for very complex dynamic systems, the integration of several control methods can fully use their respective characteristics and complementary advantages to achieve the far higher control performance than single usage of them. This viewpoint will also hold water for the unmanned underground electric locomotives because the locomotive control mathematical model is very complex and one single control method cannot regulate various system interference and uncertainties. Once efficient combination of different control methods is developed, the difficulties in control of the locomotive speed, loading and unloading, parking, emergency braking, collision avoidance, control fault tolerance and so forth can be solved. High control performance will facilitate the application of the unmanned electric locomotives in underground mines.

2. The second challenge is about speed sensorless control of the driving motor.

In the vector control of electric locomotives, the speed/position detection of the driving motor rotor must be provided. In the actual industrial control practice, the speed and position of the motor rotor are usually measured by the speed encoder. But due to harsh working surroundings and complex interaction between the locomotive body and rail, the locomotive would experience extremely strong vibration. Compared with civil railway transportation, the electric locomotives in underground mines present relatively low manufacturing accuracy and aseismic resistance. Moreover, the smoothness of underground rails is much lower than civil railway. There are lots of bends and branches in the underground mines. Particularly, the electric locomotives are in the suspending state when passing the unloading stations. Thus, all these facts cause much stronger vibration on the locomotives underground than that in the civil railway. Consequently, the service life of the speed encoders is very limited. The practice indicates that⁹ due to severe vibration, the service life of the speed encoders is very short and sometimes they cannot work normally. In addition, the external sensors greatly increase the hardware cost of the control system, to say, up to one-third of the total price.⁶⁰ Hence, it is important to solve the sensor problem in the control of the unmanned electric locomotives. A promising solution to this issue is the speed sensorless control.^61–63 The speed/position of the motor rotor can be estimated by numerical calculation without a speed sensor. This sensorless control not only improves the reliability of the control system but also optimizes its structure and reduces its manufacturing cost.

Up to date, the speed sensorless control methods include the dynamic speed estimator,⁶⁰ model reference adaptive system (MRAS),⁶² Kalman filtering,⁶³ artificial intelligence observer,⁶⁴ harmonic admittance method,⁶⁵ and high-frequency injection method.⁶⁶ The dynamic speed estimator relies on the accuracy of the motor parameters, and its speed estimation accuracy is low. The MRAS is based on the back electromotive force (EMF) to estimate the motor speed but it is difficult to detect the back EMF at low speed range because the back EMF is proportional to the motor speed. So, it is suitable for medium- and high-speed ranges. Both the Kalman filtering method and artificial intelligence observer can overcome the random interference of observation object and are workable for a wide speed range. But their computational complexity is relatively high. The high-frequency injection method is not influenced by the change of motor parameters and can accurately estimate the speed at a wide speed range, even for zero speed; but its requirement on signal processing is relatively high. Nevertheless, all these methods provide good performance for the speed sensorless control in their own applications. Successful cases have been found in the speed sensorless control of motor cars,⁶⁷ metro trains,⁶⁸ alternating current (AC) motors,⁶⁹ and electric vehicles.⁷⁰ However, little work has reported the application of the speed sensorless vector control technique in the underground unmanned electric locomotives. Although there is close similarity between the underground electric locomotive and the urban light rail, motor car, and metro train, their working environment is obviously different. Harsh working environment of the underground electric locomotives makes it impossible to directly use existing speed sensorless vector control techniques. New theories and methods are urgent for underground unmanned electric locomotives. The working environment facts should be included into the existing control technology to solve the restarting control using the speed sensorless vector control and the zero frequency compensation to the driving motor stator at extremely low speed of the underground unmanned electric locomotives. That is to say, for the specific research object of the electric locomotives, how to develop the suitable speed sensorless control method and provide reliable control performance is the key to realize the unmanned electric locomotives.

Future research directions

In section “Challenges on position tracking in underground mines,” we discuss three critical issues (i.e. data collision problem, power supply problem, and efficient data transmission) in WSN-based underground position system.

For the data collision problem, it should optimize the distribution of sensor nodes. In order for this to cover the underground working area as much as possible using the least number of nodes, the GIS information of the underground area should be considered into the WSN node optimization process. The three-dimensional (3D) parameters of the underground area, provided by GIS, will be beneficial to the distribution optimization of sensor nodes. In addition, since the locomotive is moving underground, the position system should be designed to cooperate with the locomotive movement. The UWB can be utilized with WSN to locate moving objects.⁷¹ Hence, interesting research points include the following:

GIS assisted sensor node optimization for WSN-based underground position.

Moving target locating by combination of UWB and WSN.

Precision analysis of UWB-WSN in underground position system.

In order to solve the power supply problem in WSN-based underground position, the following suggestions are made:

Development of RF-recharged batteries, including lithium-ion, supercapacitor, thin film, and solid-state batteries with RF energy-harvesting features.

Suitable recharging strategy for RF-recharged batteries.

Furthermore, CS is a promising research direction in WSN applications. The data transmission efficiency can be improved greatly between WSN nodes by CS sampling. To this end, the following future researches can be implemented.

Theoretical and experimental analyses of suitable denoiser for D-AMP applied to practical signals.

Establishment of D-AMP-based WSN for underground position.

Figure 1 depicts the future research emphases on the position tracking using underground position system.

Figure 1.

The future research directions for locomotive position tracking using underground position system.

In section “Challenges in reliable controller design,” two major challenges should be solved in the controller design of underground unmanned electric locomotive systems. One is the mathematical model about the interaction between the underground rail and locomotive body and the other is the speed sensorless control. The future research directions of these two challenges are discussed as follows.

For the mathematical model issue, it is important to first investigate the dynamics and vibrations of the locomotive body. Then, the couple effect between the locomotive body and the driving motor should be evaluated to obtain the boundary parameters and boundary conditions about the mathematical model of the locomotive body–driving motor interaction. Finally, the decoupling control of the locomotive body motion and driving motor motion needs to be developed based on the hybrid intelligent control theories. These three research points can be described as follows:

Influence mechanism of the underground rail on the dynamics of the locomotive body.

Mathematical model of the locomotive body–driving motor interaction.

MIMO controller design using hybrid intelligent control theories.

For the speed sensorless control issue, it is important to propose highly efficient speed estimation algorithm. The high-frequency injection method is very suitable for the unmanned electric locomotives because its performance is not limited by the working speed range of the driving motor. Once the motor speed can be accurately estimated, it is possible to use the high-frequency injection method to solve the restarting problem in the sensorless control. So, the research potentials include the following points:

Speed estimation algorithm using high-frequency injection method.

Restarting control under zero speed or extremely low frequency of the motor rotor.

Novel experimental device and/or experimental tests to evaluate the proposed speed sensorless control.

Figure 2 summarizes the future research emphases on the optimal control of underground unmanned electric locomotives.

Figure 2.

The future research directions for optimal control of underground unmanned electric locomotives.

Conclusion

Unmanned electric locomotive system is a critical part for smart mining. This work reviewed the progress in applications of the unmanned electric locomotive systems. The challenging tasks for locomotive position tracking and control were discussed in underground mines, and potential research directions were summarized. Through the discussion on the recent development of underground electric locomotives, it should be noticed that a comprehensive understanding about the working environment of the locomotives should be taken into account for the control system design. On one hand, WSN is a realistic solution to underground position system. The critical issues that limit the application of WSN in underground mines are discussed. In order to develop suitable WSN-based underground position system, the installation, energy consumption, and data transmission should be designed compatibly to the underground environment. The purpose of developing efficient and robust WSN is to provide accurate and real-time position information of locomotives. The position information will be used to schedule and control the movements of locomotives. On the other hand, we suggest considering the dynamics of locomotive body and underground rails into the design and manufacture of the unmanned electric locomotives. Then, based on the intelligent control theories, an advanced control system for unmanned electric locomotive systems can be developed. It is not an easy task to achieve highly efficient and intelligent control for unmanned electric locomotive systems due to that the controller design involves the computer technology, communication technology, control technology, risk assessment technology, and so forth. Once the control system is developed, the decoupling control of the locomotive body and the driving motor can be realized to provide safe, reliable, and efficient operation of unmanned electric locomotive systems.

Footnotes

Academic Editor: Hamid Reza Shaker

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research,authorship,and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research,authorship,and/or publication of this article: This research was funded by the National Natural Science Foundation of China (NSFC) (nos U1610109,51505475),the Fundamental Research Funds for the Central Universities (no. 2015XKMS018),and Priority Academic Program Development of Jiangsu Higher Education Institutions.

References

. The research on sustainable development of coal industry based on system dynamic. PhD Dissertation, Wuhan University of Technology, Wuhan, China, 2012.

Chen

. China resource efficiency: economics and outlook. Bangkok, Thailand: United Nations Environment Programme, 2013.

Mining industry is an important pillar industry in many countries of China and ASEAN. Minister of Land and Resources, http://www.chinaminingtj.org/en/index.php/en-media/352-asean-an-important-overseas-investment-destination (accessed 17 May 2013).

The statistics of safety production accidents in coal mine, State Administration of Work Safety, http://www.chinaminingtj.org/en/index.php/en-media/352-asean-an-important-overseas-investment-destination (accessed 3 June 2013)

83 people died in Shanxi coal mine safety accidents in 2012. China News Network, 14January2013, http://www.chinanews.com/df/2013/01-14/4486830.shtml

Jiang

Wang

. Multi-mode separation and nonlinear feature extraction of hybrid gear failures in coal cutters using adaptive nonstationary vibration analysis. Nonlin Dynam 2016; 84: 295–310.

Zhao

. Research on the key technology of fully mechanized coal face safety production of virtual reality system. PhD Dissertation, Xi’an University of Science and Technology, Xi’an, China, 2012.

Jiang

Hua

. A new compound faults detection method for rolling bearings based on empirical wavelet transform and chaotic oscillator. Chaos Soliton Fract 2016; 89: 8–19.

Zhang

Wang

Huang

. Research on the key technical scheme on transportation for unmanned underground mine electrical locomotive. Met Min 2013; 5: 117–120.

10.

Yang

. Research on unmanned mine car control system in metal mines. MS Thesis, Kunming University of Science and Technology, Kunming, China, 2012.

11.

Jiang

Hua

. The nonlinear dynamics response of cracked gear system in a coal cutter taking environmental multi-frequency excitation forces into consideration. Nonlin Dynam 2016; 84: 203–222.

12.

Pukkila

Sarrka

. Intelligent mine technology program and its implementation. In: Proceedings of MassMin 2000, Brisbane, QLD, Australia, 29 October–2 November 2000. Carlton, VIC, Australia: Australasian Institute of Mining and Metallurgy.

13.

Gumanyuk

Zakharchuk

Lisovets

. An automatic system for controlling and regulating the speed of an electric locomotive. Mekhanizahriya Avtom Upravl 1969; 4: 18–21.

14.

Wang

. Development trend and science technology development strategy of mining technology for metallic mines. Met Min 2006; 1: 19–25.

15.

Gaspari

. Armchair mining. Mech Eng 2003; 125: 42.

16.

Seoble

Daneshmend

. Mine of the year 2020: technology and human resources. CIM Bull 1998; 91: 51–60.

17.

Copco

. Good results with teleremote. Min Constr 2003; 2: 6–7.

18.

Mining robotics project, CSIRO, https://research.csiro.au/robotics/mine-informatics/ (accessed 1 June 2011)

19.

Pukkila

. Implementation of mine automation: the importance of work safety and motivation. Acta polytechnica Scandinavica-Civil engineering and building construction series, Helsinki: The Finnish Academy of Technology, 1999.

20.

1998 Annual meeting presentations, http://www.vale.com/canada/EN/Pages/default.aspx

21.

Zhang

. Rio invests 5.18 billion in the automation of the Unmanned Mine Train, http://www.smh.com.au/business/mining-and-resources/rio-tinto-pushes-ahead-with-driverless-trains-in-pilbara-20150309-13z28v.html

22.

ENFI. The technical of unmanned underground mine electrical locomotive. Mod Occup Saf 2013; 4: 108.

23.

Macfarlane

. The implementation of new technology in southern African mines: pain or panacea. J South Afr Inst Min Metall 2001; 101: 115–126.

24.

Golde

. Implementation of drill teleoperation in mine automation. Montreal, Quebec, Canada: McGill University, 1997.

25.

Hofmann-Wellenhof

Lichtenegger

Collins

. Global positioning system: theory and practice. New York, NY: Springer Science & Business Media, 2012.

26.

Misra

Enge

. Global positioning system: signals, measurements and performance. 2nd ed.Lincoln, MA: Ganga-Jamuna Press, 2006.

27.

Farrell

. Aided navigation: GPS with high rate sensors. McGraw-Hill, Inc, 2008.

28.

Cypriani

Delisle

Hakem

. Wi-fi-based positioning in a complex underground environment. J Netw 2015; 10: 141–151.

29.

Infrasurvey Sàrl, http://www.infrasurvey.ch/geomon/?lang=en

30.

Finnsson

. The DigiTrak Eclipse locating system: a flexible and innovative approach to meeting today’s HDD locating challenges. In: Proceedings of 23rd no-dig international conference and exhibition 2005, Rotterdam, 19–21 September 2005, pp.161–170. London: International Society for Trenchless Technology.

31.

Nagovitsyn

. Accuracy of location systems for horizontal directional drilling (HDD). In: Proceedings of 26th no-dig international conference and exhibition 2008, Moscow, Russia, 3–6 June 2008, pp.108–111. London: International Society for Trenchless Technology.

32.

Chen

Zhou

. RFID-based underground personnel positioning system design. Adv Mat Res 2014; 850–851: 653–656.

33.

Maqsud

Daku

Prugger

. Characterization of a MEMS accelerometer for an underground mine positioning system. In: Proceedings of 2005 Canadian conference on electrical and computer engineering (Chair R Billinton), Saskatoon, SK, Canada, 1–4 May 2005, pp.2268–2271. New York, NY: IEEE.

34.

Conceicao

Pendao

Moreira

. Evaluation of medium access and a positioning system in wireless underground sensor networks. In: Proceedings of 2016 wireless days, Toulouse, France, 23–25 March 2016, pp.1–6. New York, NY: IEEE.

35.

Liu

. Design of underground precise positioning service system based on CSS. Ind Min Autom 2015; 41: 40–44.

36.

Salih

Despins

Affes

. Ultra-wideband fast acquisition system for positioning in an underground mining environment. In: Proceedings of 2008 international conference on communications, Beijing, China, 19–23 May 2008, pp.1389–1395. New York, NY: IEEE.

37.

Liu

Wang

. Underground personnel positioning system based on ultra wideband technology. Ind Min Autom 2014; 40: 81–83.

38.

Katayama

. An experimental trial of location and route guidance in underground mall by GPS repeater system. J Inst Electron Inform Commun Eng 2009; 92: 276–280.

39.

Chen

Xiao

. Study on signal conflict of underground personnel positioning system of based on ZigBee. In: Proceedings of 2nd international conference on artificial intelligence, management science and electronic commerce, Dengleng, China, 8–10 August 2011, pp.659–661. New York, NY: IEEE.

40.

Baffour

Abatan

. Developing an underground infrastructure management system using GPR, GPS and GIS. In: Proceedings of the IASTED international conference information and knowledge sharing, St. Thomas, VI, 18–20 November 2002, pp.163–166. Calgary, AB, Canada: Acta Press.

41.

Kang

Xie

Huang

. Compressive sensing technology in distributed wireless sensor networks. Signal Process 2013; 29: 1560–1567.

42.

Liu

Xing

. Key issues analysis and experiment of the global positioning system in underground tunnel network. J China Coal Soc 2011; 36: 519–526.

43.

Wang

Niyato

. Wireless networks with RF energy harvesting: a contemporary survey. IEEE Commun Surv Tut 2015; 17: 757–789.

44.

Visser

Vullers

. RF energy harvesting and transport for wireless sensor network applications: principles and requirements. Proc IEEE 2013; 101: 1410–1423.

45.

Donoho

. Compressed sensing. IEEE T Inform Theory 2006; 52: 1289–1306.

46.

Zhang

Zhao

. Image compressive sensing recovery using adaptively learned sparsifying basis via L0 minimization. Signal Process 2014; 103: 114–126.

47.

Haupt

Nowak

. Signal reconstruction from noisy random projections. IEEE T Inform Theory 2006; 52: 4036–4048.

48.

Block

Uecker

Frahm

. Undersampled radial MRI with multiple coils. Iterative image reconstruction using a total variation constraint. Magn Reson Med 2007; 57: 1086–1098.

49.

Eldar

Kuppinger

Bolcskei

. Block-sparse signals: uncertainty relations and efficient recovery. IEEE T Signal Proces 2010; 58: 3042–3054.

50.

Baraniuk

Cevher

Duarte

. Model-based compressive sensing. IEEE T Inform Theory 2010; 56: 1982–2001.

51.

Som

Schniter

. Compressive imaging using approximate message passing and a Markov-tree prior. IEEE T Signal Proces 2012; 60: 3439–3448.

52.

Dong

Shi

. Compressive sensing via nonlocal low-rank regularization. IEEE T Image Process 2014; 23: 3618–3632.

53.

Zhang

Rao

. Extension of SBL algorithms for the recovery of block sparse signals with intra-block correlation. IEEE T Signal Proces 2013; 61: 2009–2015.

54.

Metzler

Maleki

Baraniuk

. From denoising to compressed sensing. IEEE T Inform Theory 2016; 62: 5117–5144.

55.

Yan

Qin

. Model reference robust control for marine propulsion systems with model uncertainty caused by hull deformation. J Mar Sci Technol 2013; 21: 400–409.

56.

Wang

. Research on speed sensorless indirect stator quantities control of traction motor. PhD Dissertation, Central South University, Changsha, China, 2010.

57.

Shao

Dong

. Control strategy review of permanent magnet synchronous motor servo system. Elect Autom 2013; 35: 1–4.

58.

Yan

Qin

. Robust global sliding model control for water-hull-propulsion unit interaction systems—part 1: system boundary identification. Teh Vjesn 2015; 22: 209–215.

59.

Yan

Qin

. Robust global sliding model control for water-hull-propulsion unit interaction systems—part 2: model validation. Teh Vjesn 2015; 22: 465–473.

60.

Tang

Yang

. MRAS based speed sensorless flywheel battery drive for charging procedure. Smal Spec Electr Mach 2010; 6: 58–61.

61.

Tajima

Hori

. Speed sensorless field-orientation control of the induction machine. IEEE Trans Ind App 1993; 29: 175–180.

62.

Wang

Zhang

Liu

. Sensorless of integral sliding mode controlled PMSM at low speeds. Electr Mach Cont 2012; 12: 19–26.

63.

Chen

Liu

Jin

. Direct torque control of permanent magnet synchronous motors based on extended Kalman filter observer of flux linkage. Proc CSEE 2012; 12: 19–26.

64.

Huang

Kuang

. PMSM sensorless control based on fuzzy active-disturbance rejection. J Hunan Univ 2012; 39: 37–44.

65.

Yao

Tian

Zhang

. Novel method for motor’s sensorless speed detection. Electr Dr 2007; 37: 57–61.

66.

Zhu

. Rotating high frequency signal injection method based on FIR optimization filter and its application. Chin J Sci Inst 2012; 33: 588–595.

67.

Ding

Gui

. Research of speed sensorless control technology and its application in high power AC traction drive. J China Railw Soc 2008; 30: 75–84.

68.

Shang

Liu

Nian

. Speed sensorless control and restarting at unknown speed of traction motor. Proc CSEE 2006; 26: 118–123.

69.

Jiang

Sun

. Development and application of speed sensorless vector control technology. Foreign Electron Meas Technol 2007; 26: 71–74.

70.

Liu

. Research on speed sensorless vector control system of EV. Mech Eng Autom 2013; 4: 117–121.

71.

Sjogren

Pettersson

. Detection of moving targets by focusing in UWB SAR—theory and experimental results. IEEE T Geosci Remote 2010; 48: 3799–3815.

Recent progress on smart mining in China: Unmanned electric locomotive

Abstract

Keywords

Introduction

Progress on unmanned electric locomotive

Challenges on position tracking in underground mines

Challenges in reliable controller design

Future research directions

Conclusion

Footnotes

Declaration of conflicting interests

Funding

References