WO2022088316A1 - 一种列车动力分配方法及装置 - Google Patents

一种列车动力分配方法及装置 Download PDF

Info

Publication number
WO2022088316A1
WO2022088316A1 PCT/CN2020/130343 CN2020130343W WO2022088316A1 WO 2022088316 A1 WO2022088316 A1 WO 2022088316A1 CN 2020130343 W CN2020130343 W CN 2020130343W WO 2022088316 A1 WO2022088316 A1 WO 2022088316A1
Authority
WO
WIPO (PCT)
Prior art keywords
power
train
vehicle
motor
state
Prior art date
Application number
PCT/CN2020/130343
Other languages
English (en)
French (fr)
Inventor
刘可安
赵旭峰
尚敬
徐绍龙
甘韦韦
郭维
吴业庆
李科
喻励志
王亮
Original Assignee
株洲中车时代电气股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株洲中车时代电气股份有限公司 filed Critical 株洲中车时代电气股份有限公司
Priority to AU2020474272A priority Critical patent/AU2020474272A1/en
Publication of WO2022088316A1 publication Critical patent/WO2022088316A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B61—RAILWAYS
    • B61C—LOCOMOTIVES; MOTOR RAILCARS
    • B61C15/00—Maintaining or augmenting the starting or braking power by auxiliary devices and measures; Preventing wheel slippage; Controlling distribution of tractive effort between driving wheels
    • B61C15/14—Maintaining or augmenting the starting or braking power by auxiliary devices and measures; Preventing wheel slippage; Controlling distribution of tractive effort between driving wheels controlling distribution of tractive effort between driving wheels

Definitions

  • the invention relates to a power distribution technology of a wheel-rail type train, in particular to a train power distribution method and a train power distribution device.
  • Traction and electric braking force control is one of the core functions of train control, and its control performance directly affects the safety and reliability of train operation.
  • the current power distribution of multi-power unit trains usually adopts the method of synchronous and equal sharing, that is, the driver gives the total required power of the train through the control handle, and then the train control system distributes the total power of the train according to the method of equal distribution.
  • Demand power delivery is a given torque for each power unit.
  • the power exertion of the train is limited by the physical adhesion coefficient between the wheel and rail.
  • the power distribution of the train is more distributed, and the wheelset self-cleaning effect of the previous power unit will improve the track surface condition of the subsequent power unit. Therefore, the wheel-rail adhesion conditions where each power unit is located will show great differences. Under such conditions, the existing synchronous equal-split power setting mode will not only increase the idling frequency of locomotives with poor adhesion conditions and reduce the smoothness of its traction performance, but also limit the traction performance of locomotives with better adhesion conditions.
  • the present invention provides a train power distribution method, a train power distribution device, and a computer-readable storage medium, which are used to analyze the trains of each train according to the dynamic state of each train.
  • the power distribution of the train is asynchronously coordinated and optimized, so as to realize the multi-objective traction optimization control such as the maximum traction performance of the train, the minimum longitudinal impact, and the optimal converter system state.
  • the train includes a multi-carriage vehicle.
  • the multi-section vehicle is divided into a motor car and a trailer.
  • the train power distribution method includes: according to the travel planning curve of the train, the travel route information and the vehicle state of each of the motor cars, with the goal of minimum longitudinal impact, allocating the total power of the train to each of the motor cars; the state of a plurality of power units of the motor vehicle, and aiming at the optimal inverter system state, the given power distributed to the motor vehicle is further distributed to each of the power units of the motor vehicle; and according to the power unit of each power unit The wheel-rail adhesion state maximizes the execution of the given power distributed to the power unit.
  • the travel planning curve may include a planning speed curve and a planning power curve, which are used to indicate the train speed and the total power of the train at each moment of the travel route.
  • the travel route information may include a slope gradient and a curve radius of the train at the current moment.
  • the vehicle state of each of the motor vehicles may include the maximum allowable power of the motor vehicle fed back by each of the motor vehicles.
  • the step of allocating the total power of the train to each of the high-speed trains may include: taking the total train power distribution scheme as a solution object, and according to the driving planning curve, the driving route information and the vehicle state of each of the high-speed trains, for all the high-speed trains.
  • Quantitative modeling is performed on the longitudinal impact of the train, wherein the total power distribution scheme of the train indicates the power distributed to each of the moving trains; and the constructed longitudinal impact quantification model is optimized and solved to obtain the train corresponding to the minimum longitudinal impact. Total power distribution scheme.
  • the step of quantitatively modeling the longitudinal impact of the train may include: according to the travel planning curve, the travel route information, and the vehicle state of each motor train, calculating coupling force and coupling force impulse between the vehicles; and quantifying the longitudinal impact of the train according to the maximum coupler force and the maximum coupler force impulse between the vehicles to construct the longitudinal impact quantification model.
  • the step of optimally solving the longitudinal impact quantification model may include: adopting a control variable parameterization method or a swarm intelligence algorithm, when the single-cycle power of each motor vehicle is the largest
  • the optimal solution for the total power distribution scheme of the train is carried out within the range of the allowable variation, wherein the maximum allowable variation of the single-cycle power of the motor vehicle is determined by the vehicle speed of the motor vehicle and/or the state of the converter system.
  • the train power distribution method may further include: first limiting the total train power at each moment of the planned power curve according to the maximum allowable variation of the power of the train in a single cycle. amplitude filtering processing, wherein the maximum allowable variation of the power of the train in a single cycle is determined by the train speed, train network pressure and/or driving line conditions; The longitudinal shock of the train is quantitatively modeled.
  • the train power distribution method may further include: calculating the maximum allowable power of the train at the corresponding moment according to the maximum allowable power of each of the moving trains; in response to the maximum allowable power of the train being less than For the total train power at the corresponding moment of the planned power curve processed by the limiting filtering, the train power distribution scheme consisting of the maximum allowable power of each motor car is substituted into the longitudinal shock quantification model to calculate the corresponding train shock quantification value; in response to the train impact quantization value being less than a quantization threshold value, the total power of the train is distributed according to the maximum allowable power of each of the moving trains, wherein the quantization threshold value is the maximum allowable longitudinal direction obtained according to the train operation safety assessment determining a shock quantization value; and in response to the train shock quantization value being greater than or equal to the quantization threshold value, gradually reducing the total train power at the corresponding moment until the train shock quantization value is smaller than the quantization threshold value.
  • the vehicle state of the motor vehicle may include the wheel-rail adhesion state coefficient fed back by each of the power units of the motor vehicle, the maximum allowable power of the unit, the comprehensive motor speed and/or the motor comprehensive speed. temperature.
  • the step of further distributing the given power distributed to the own motor vehicle to each of the power units of the own motor vehicle may include: in response to the motor vehicle maximum allowable power being greater than or equal to the given power distributed to the own motor vehicle, distributing the power to the motor vehicle.
  • the scheme is the solution object, and the state of the converter system is quantitatively modeled according to the vehicle state of the motor vehicle.
  • the power distribution scheme of the motor vehicle indicates the power of each of the power units distributed to the motor vehicle, and the motor vehicle is the largest
  • the allowable power is calculated according to the unit maximum allowable power of each of the power units; and within the range of the single-cycle maximum allowable power variation of each of the power units of the motor vehicle, the constructed system state quantification model is optimized and solved,
  • the maximum allowable variation of the single-cycle power of the power unit is determined by the vehicle speed of the motor vehicle and/or the inverter system state.
  • the step of further distributing the given power allocated to the own motor vehicle to each of the power units of the own motor vehicle may further include: in response to the maximum allowable power of the motor vehicle being less than the power allocated to the own motor vehicle
  • the step of further distributing the given power allocated to the own motor vehicle may further include: in response to the maximum allowable power of the motor vehicle being less than the power allocated to the own motor vehicle
  • For the given power of the motor car take the power distribution scheme of the motor car as the solution object, according to the power exertion objective function of the motor car and the converter state objective function, the power exertion of the motor car and the state of the converter system are comprehensively and quantitatively modeled.
  • the power exertion objective function indicates the sum of the powers allocated to the power units of the own motor vehicle
  • the converter state objective function indicates the quantified value of the state of the converter system
  • the step of further distributing the given power distributed to the host vehicle to each of the power units of the host vehicle may further include: in response to the given power distributed to the host vehicle being less than The power threshold value centrally distributes the given power allocated to the own motor vehicle to some power units of the own motor vehicle, wherein the power threshold value is determined according to the energy efficiency of each of the power units of the own motor vehicle.
  • the step of maximizing the execution of the given power distributed to the power unit may include: calculating the wheel set of the power unit according to the creep speed of the wheel set and the acceleration index of the wheel set Rail adhesion state coefficient; in response to the wheel-rail adhesion state coefficient indicating that the wheelset has no idling sliding tendency, the given power distributed to the power unit is sent to the inverter controller as the adhesion given force; in response to the The wheel-rail adhesion state coefficient indicates that the wheelset has a tendency of idling or has occurred, and the wheelset adhesion is calculated according to the effective wheel diameter of the power unit, the moment of inertia of the wheelset, the speed of the wheelset and the actual force exerted by the motor, and sending the wheelset adhesion force to the inverter controller as an adhesion given force; and using the inverter controller to control the traction motor of the power unit to execute the adhesion given force.
  • the train power distribution method may further include: feeding back the wheel-rail adhesion state coefficient of the power unit to the vehicle-level controller of the corresponding motor car, so as to form the Corresponding to the vehicle state of the motor vehicle; in response to the wheel-rail adhesion state coefficient indicating that the wheelset has no idling sliding tendency, according to the maximum allowable power at the current speed of the shaft end of the power unit, the given power at the previous moment and the wheel
  • the rail adhesion state coefficient calculates the wheelset adhesion force at the current time, and feeds the wheelset adhesion force at the current time back to the vehicle-level controller for composing the vehicle state of the corresponding motor vehicle; and in response to the The wheel-rail adhesion state coefficient indicates that the wheelset has a tendency to idling or has occurred, and the maximum value of the adhesion of the wheelset in a complete idling control cycle is selected and fed back to the vehicle-level controller for the composition
  • the vehicle-rail adhesion state coefficient indicates
  • the step of maximizing the execution of the given power distributed to the power unit may further include: collecting the temperature, current, voltage and rotational speed of the traction motor of the power unit to determine the traction inverter state of the power unit; in response to the traction inverter state being good, controlling the traction motor to execute the sticking given force; and in response to the traction inverter state being poor, according to the traction inverter state
  • the state limits the power of the traction motor and calculates the corresponding limited power, and compares the sticking given force with the limited power to control the traction motor to execute the smaller value.
  • the power unit may comprise at least one control shaft.
  • Each control shaft corresponds to at least one traction motor.
  • the power distribution method may further include: taking an average of the rotational speeds of the traction motors as a comprehensive motor rotational speed of the power unit, and feeding back the motor comprehensive rotational speed to a vehicle-level controller of the corresponding motor vehicle for use in the motor vehicle.
  • the vehicle state of the corresponding motor vehicle is formed; the temperature of each traction motor is averaged as the motor comprehensive temperature of the power unit, and the motor comprehensive temperature is fed back to the vehicle-level controller for forming the vehicle state corresponding to the motor vehicle; and taking the smaller value of the sticking given force and the limit power as the unit maximum allowable power of the power unit, and feeding back the unit maximum allowable power to the power unit A vehicle-level controller for composing the vehicle state of the corresponding motor vehicle.
  • the train power distribution method may further include: obtaining the travel planning curve from an automatic driving system of the train; obtaining the travel route information from a train operation monitoring and recording device ; and obtain the vehicle state of each of the motor cars from the vehicle-level controller of each of the motor cars, wherein each of the vehicle states is formed by a combination of unit states fed back to the vehicle-level controller by a plurality of power units corresponding to the motor car, respectively. become.
  • a train power distribution device is also provided herein.
  • the train includes a multi-carriage vehicle.
  • the multi-section vehicle is divided into a motor car and a trailer.
  • the train power distribution device includes a memory and a processor.
  • the processor is connected to the memory and is configured to implement the train power distribution method provided by any one of the above embodiments.
  • a computer-readable storage medium is also provided herein.
  • the above-mentioned computer-readable storage medium provided by the present invention stores computer instructions thereon.
  • the train power distribution method provided by any one of the above embodiments can be implemented.
  • FIG. 1 shows a schematic flowchart of a train power distribution method according to some embodiments of the present invention.
  • FIG. 2 shows a schematic diagram of longitudinal force on a train according to some embodiments of the present invention.
  • FIG. 3 shows a schematic diagram of an optimal set front according to some embodiments of the present invention.
  • FIG. 4 shows a schematic diagram of optimal adhesion points provided in accordance with some embodiments of the present invention.
  • the present invention provides a train power distribution method, a train power distribution device, and a computer-readable storage medium. Based on the comprehensive constraints of the dynamic state, wheel-rail state and converter state of each power unit of the train, the present invention can carry out the train power (here the power includes the traction force and the electric braking force, and the power is referred to as the traction force and the electric braking force in the following text). Power) asynchronous collaborative optimization, so as to achieve multi-objective traction optimization control such as maximum traction performance, minimum longitudinal impact, and optimal converter system state of autonomous trains.
  • the above-described train power distribution method may be automatically implemented by a train power distribution device.
  • the train power distribution device may include a memory and a processor.
  • the memory may include a computer-readable storage medium having computer instructions stored thereon.
  • the processor may be coupled to the memory and configured to execute computer instructions stored on the memory to implement the train power distribution method described above.
  • FIG. 1 shows a schematic flowchart of a train power distribution method according to some embodiments of the present invention.
  • the above-mentioned train power distribution method provided by the present invention may include the steps of: performing train-level power coordinated control according to the train's travel planning curve, travel route information and vehicle status of each motor train.
  • the above-mentioned train-level power cooperative control may be implemented by a train-level power cooperative controller.
  • the train-level power cooperative controller can be realized by the processor of the train power distribution device, and is at the first control level of the train with the main engine of the automatic driving system of the train. It is mainly used for planning curves inputted by the automatic driving system and train operation monitoring records.
  • the line information input by the device (LKJ) and the vehicle state fed back by each motor car can predict the longitudinal dynamic state of the train, so as to minimize the longitudinal impact as the goal, the total power of the train is optimally distributed to each motor car.
  • the train-level power cooperative controller can first obtain its planned travel planning curve from the automatic driving system of the train, obtain the current travel route information from the train operation monitoring and recording device (LKJ), and obtain the travel route information from the vehicle-level of each motor train.
  • the controller obtains the vehicle state of each motor car as the input information of the train-level power cooperative control.
  • the planned speed curve V train may include a plurality of elements, each element indicating the train speed of the train at a corresponding moment of the travel route.
  • the planned power curve F train may include a plurality of elements, each of which indicates the total train power of the train at a corresponding moment of the running route.
  • the travel route information may include the ramp gradient i, the ramp length l i , the curve radius R and the curve length l r of the train at the current moment.
  • the maximum allowable power F limit of the motor car includes a plurality of elements, and each element indicates the maximum allowable power of a motor car (for example: F M1_limit ).
  • the maximum allowable power F M1_limit of the one motor car can be calculated by feeding back the multiple power units of the motor car M1 to the sum of the maximum allowable power F uj_limit of multiple units of its vehicle-level controller.
  • the train-level power cooperative controller can execute the train-level power cooperative control according to the above input information, and distribute the total train power asynchronously and cooperatively to each EMU of the train, so as to solve the problem of the existing locomotive power synchronous distribution method in large marshalling, complex curves, etc. Under severe working conditions, it is easy to cause the problem of limited power exertion.
  • the train-level power cooperative controller may first perform limit filtering on the received planned power curve F train :
  • f dec is the maximum allowable drop of the train power in a single cycle, which is determined by the performance of the train controlled by the system.
  • the output value of the relevant function; fris is the maximum allowable ascent of the train power in a single cycle, which is determined by the performance of the system to control the train. It can be a fixed value or a function output value related to real-time status such as train speed, train network pressure, and line conditions. .
  • FIG. 2 shows a schematic diagram of longitudinal force on a train according to some embodiments of the present invention.
  • the train may include multiple vehicles, which are divided into motor cars and trailers.
  • M is the identification of the motor vehicle
  • T is the identification of the trailer.
  • the motor vehicle and the trailer can be numbered in the form of consecutive numbers (ie, M1, T2...Tn-1, Mn, Tn+1).
  • the function min J[F set (t)] is used to calculate the The total train power distribution scheme F set that obtains the minimum value.
  • the total train power distribution scheme F set includes a plurality of elements, each of which indicates the power distributed to one motor car.
  • M k (a+bv+cv 2 ) is the running resistance of the k vehicle, where M k is the mass of the k vehicle; the drag coefficient (a M , b M , c M ) of the powered vehicle is obtained by offline identification; the trailer adopts the average trailer
  • the drag coefficients (a T , b T , c T ) are identified online by the recursive least squares method.
  • a Mk (t) is the instantaneous acceleration of the powered vehicle at time t, which is obtained by numerical processing of the axle speed sensor signal.
  • a Tk (t) is the reference acceleration of the trailer.
  • a Tk (t) may be obtained by numerical processing of the trailer's speed sensor signal.
  • a Tk (t) can be obtained by using the train reference acceleration through numerical differentiation processing from the train reference speed.
  • M Tall is the total trailer mass, that is, the traction load of the train.
  • ⁇ is the identification parameter of the recursive least squares online parameter identification
  • ⁇ is the current state quantity of the train.
  • the above is based on the maximum coupler force and maximum coupler force impulse
  • the scheme of constructing the longitudinal impact quantification model is only a non-limiting case provided by the present invention, which is intended to clearly demonstrate the main idea of the present invention, and provide a specific scheme that is convenient for the public to implement, but is not intended to limit the present invention. the scope of protection of the invention.
  • those skilled in the art can also use indicators such as speed, acceleration, and acceleration differential to replace the above-mentioned maximum coupler force based on the concept of the present invention. and maximum coupler force impulse
  • the effect of quantifying the longitudinal shock of the train can also be achieved.
  • the train-level dynamic cooperative controller can use the control variable parameterization method (CVP) or the swarm intelligence algorithm (for example: PSO particle swarm algorithm, TLBO teaching and learning algorithm), in the train total
  • CVP control variable parameterization method
  • the swarm intelligence algorithm for example: PSO particle swarm algorithm, TLBO teaching and learning algorithm
  • the feasible solution domain of the total train power distribution scheme F set is F k_set (t) ⁇ F k_limit (t)&F k_set (t-1)-f vk_dec ⁇ F k_set (t) ⁇ F k_set (t-1)+f vk_ris , where k is the number of the power vehicle; f vk_dec is the maximum allowable drop of power in a single cycle of the power vehicle k, which is determined by the vehicle traction conversion performance and the train impact rate limit, and can be a fixed value or a Fvk_ris is the output value of the function related to the real-time state such as the state of the power train; f vk_ris is the maximum allowable rise of the power vehicle k in a single cycle, which is determined by the traction conversion performance of the vehicle and the limit of the train impact rate. It can be a fixed value or a value related to the vehicle speed, Function output value related to real-time status such as converter status.
  • the train-level power cooperative controller may further calculate the maximum allowable power of the train at the corresponding moment according to the maximum allowable power of each motor car, and process the maximum allowable power of the train and the limit filter after processing.
  • the total power of the train at the corresponding moment of the planned power curve Make a comparison to judge the size of the two. like Then the train-level power cooperative controller can judge the total power of the train given by the automatic driving system. If the limit of the maximum allowable power of the train is not exceeded, the above-mentioned total power distribution scheme F set of the train can be realized.
  • the train-level power cooperative controller can judge the total power of the train given by the automatic driving system. If the limit of the maximum allowable power of the train is exceeded, the above-mentioned total power distribution scheme F set of the train cannot be realized. At this time, the train-level power cooperative controller needs to substitute the train power distribution scheme F limit composed of the maximum allowable power of each motor car into the above-mentioned longitudinal shock quantification model to calculate the corresponding train shock quantization value J[F limit (t)]. After that, the train-level power cooperative controller may determine the magnitude of the train impact quantization value J[F limit (t)] and the preset quantization threshold value J limit .
  • the quantified threshold value J limit can be determined according to the quantified value of the maximum allowable longitudinal impact obtained from the train operation safety assessment, and is mainly used to ensure the safe operation of the train. If J[F limit (t)] ⁇ J limit , it means that the train power distribution scheme F limit meets the requirements of train operation safety assessment, and the train-level power cooperative controller can allocate the total power of the train according to the maximum allowable power F limit of each motor car. .
  • the train-level power cooperative controller needs to gradually reduce the step size according to a preset fixed increment. until At this time, the train-level power cooperative controller can obtain a feasible solution, so that the train impact quantization value is less than the quantization threshold value, that is, J[F limit (t)] ⁇ J limit .
  • the train-level power cooperative controller may reduce the Feedback to the train's autopilot system, so that the autopilot system will Real-time correction of power setting curve
  • the vehicle-level power distribution control described above may be implemented by a vehicle-level power distribution controller.
  • the vehicle-level power distribution controller can be implemented by the processor of the train power distribution device, and is configured in the vehicle network or logic control level (ie, the second control level) of each motor car.
  • the vehicle-level power distribution controller is mainly used to target the optimal converter system state according to the wheel-rail adhesion state of each power unit, the efficiency and temperature rise of each power traction converter system, and the health state of each power unit.
  • the given power allocated by the train-level power coordination layer is optimally distributed to each power unit of the EMU.
  • the vehicle-level power distribution controller may first obtain the power unit states fed back by each power unit of the own motor vehicle to form the vehicle state of the own motor vehicle.
  • the wheel-rail adhesion state coefficient ⁇ includes a plurality of elements, each of which indicates a wheel-rail adhesion state coefficient of a power unit of the motor vehicle.
  • the unit maximum allowable power Fu_limit includes a plurality of elements, each element indicating the unit maximum allowable power of a power unit of the host vehicle.
  • the comprehensive motor speed ⁇ includes a plurality of elements, and each element indicates the motor comprehensive speed of a power unit of the motor vehicle.
  • the motor comprehensive temperature T includes a plurality of elements, and each element indicates the motor comprehensive temperature of a power unit of the motor vehicle.
  • the vehicle-level power distribution controller may sum the elements in the unit's maximum allowable power F u_limit to calculate the own motor vehicle's maximum allowable power F Mk_limit , namely:
  • N is the number of power units of the power vehicle
  • F uj is the maximum allowable power of the unit fed back by the power unit j.
  • the motor vehicle power distribution scheme F u_set includes a plurality of elements, each element indicating a given power distributed to one power unit of the host motor vehicle.
  • the vehicle-level power distribution controller may select an appropriate distribution scheme to perform vehicle-level power distribution control according to the magnitude relationship between the given power F Mk_set and the maximum allowable power F Mk_limit of the motor vehicle.
  • the vehicle-level power distribution controller can determine that the given power F Mk_set does not exceed the maximum power of the motor vehicle With the limit of the allowable power F Mk_limit , the power vehicle can fully exert the given power F Mk_set issued by the train-level power cooperative controller. At this time, the vehicle-level power distribution controller does not need to consider the maximum exertion of power, and only takes the optimal state of the converter as the goal, and quantitatively models the state of the converter system according to the vehicle state of the motor vehicle:
  • ⁇ tol is the total vehicle efficiency
  • K eff is the adjustable weight gain of the efficiency index
  • ⁇ T tol is the total temperature rise of the system
  • K T is the adjustable weight gain of the temperature rise index
  • S tol is the system noise
  • K SIL is the adjustable weight gain of the system noise
  • ⁇ adh is the vehicle sticking state, the smaller the index is, the easier it is to idling
  • K adh is the adjustable weight gain of the vehicle sticking state.
  • K eff ⁇ tol +K T / ⁇ T tol +K SIL /S tol +K adh ⁇ adh is the quantized value of the system state of the converter, and the larger the quantized value, the better the system state of the converter.
  • the function max J[F u_set (t)] is used to calculate the motor vehicle power distribution scheme Fu_set that can maximize K eff ⁇ tol +K T / ⁇ T tol +K SIL /S tol +K adh ⁇ adh .
  • Model eff is a power unit efficiency prediction model, a mechanism model or an empirical model based on test data.
  • Model Tep is a temperature rise prediction model of the power unit, a mechanism model or an empirical model based on test data.
  • Model SIL is a noise prediction model of a power unit, a mechanism model or an empirical model based on test data.
  • the vehicle-level power distribution controller can adopt the control variable parameterization method (CVP) or the swarm intelligence algorithm (for example: PSO particle swarm algorithm, TLBO teaching and learning algorithm), in the motor vehicle dynamic
  • CVP control variable parameterization method
  • the swarm intelligence algorithm for example: PSO particle swarm algorithm, TLBO teaching and learning algorithm
  • F u_set is F uj_set (t) ⁇ F uj_limit (t)&F uj_set (t-1)-f uj_dec ⁇ F uj_set (t) ⁇ F uj_set (t-1)+f uj_ris .
  • f uj_dec is the single-cycle drop limit of the power unit j, which is determined by the performance of the power unit and the limit of the impact rate of the vehicle. It can be a fixed value or a function output value related to the real-time status such as vehicle speed and converter status.
  • f uj_ris is the single-cycle rise limit of the power unit j, which is determined by the vehicle traction conversion performance, which can be a fixed value or a function output value related to real-time status such as vehicle speed and converter status.
  • the vehicle-level power distribution controller can determine that the given power F Mk_set exceeds the limit of the maximum allowable power F Mk_limit of the motor vehicle. , the EMU cannot meet the given power F Mk_set issued by the train-level power cooperative controller. At this time, the vehicle-level power distribution controller needs to consider the two optimization objectives of the maximum power exertion and the optimal state of the converter at the same time.
  • f 1 (t) is the objective function of power exertion, indicating the sum of the given power distributed to each power unit of the motor vehicle.
  • f 2 (t) is the converter state objective function, which indicates the quantized value of the converter system state.
  • [f 1 (t), f 2 (t)] indicates the comprehensive quantization value of the power exertion and the converter state. The larger the quantization value, the better the comprehensive state.
  • the function max f(t) is used to calculate the motor vehicle power distribution scheme F u_set which can make [f 1 (t), f 2 (t)] take the maximum value.
  • the vehicle-level power distribution controller can use swarm intelligence algorithms (such as PSO particle swarm algorithm, TLBO teaching and learning algorithm) to analyze the constructed system state in the feasible solution domain of the EMU power distribution scheme F u_set
  • swarm intelligence algorithms such as PSO particle swarm algorithm, TLBO teaching and learning algorithm
  • the quantitative model is optimized to obtain the Pareto optimal set frontier corresponding to the best comprehensive state.
  • the feasible solution domain of the motor vehicle power distribution scheme F u_set is F uj_set (t) ⁇ F uj_limit (t)&F uj_set (t-1)-f uj_dec ⁇ F uj_set (t) ⁇ F uj_set (t-1) +f uj_ris .
  • FIG. 3 shows a schematic diagram of an optimal set front according to some embodiments of the present invention.
  • the Pareto optimal set frontier corresponding to the best synthesis state may include multiple data points. Each data point may indicate a motor vehicle power distribution scheme Fu_set corresponding to the best comprehensive state.
  • the vehicle-level power distribution controller can be preset according to the minimum traction limit criterion, the minimum converter state limit criterion and the priority of traction force exertion and converter state, and further from the multiple power distribution schemes of the Pareto optimal set front.
  • An optimal solution F u_set satisfying the conditions is selected to implement vehicle-level power distribution control.
  • those skilled in the art can also use other algorithms such as neural networks, deep learning and other algorithms to optimize and solve the above-mentioned comprehensive quantification model based on the concept of the present invention, and obtain the best comprehensive quantification model by the same calculation.
  • the Pareto optimal set front surface corresponding to the state is selected, and an optimal solution F u_set that satisfies the conditions is selected to implement the above-mentioned vehicle-level power distribution control.
  • the vehicle-level power distribution controller may further determine whether to perform axis cutting control to improve vehicle efficiency according to the given power F Mk_set distributed to the motor vehicle and a preset power threshold value F Mk_th .
  • the power threshold value F Mk_th may be determined by the energy efficiency of each power unit of the own motor vehicle, and is used to indicate the sum of the minimum powers that can make the power units of the own motor vehicle operate efficiently.
  • F Mk_set ⁇ F Mk_th , it means that the given power allocated to the motor vehicle is relatively small, and the vehicle-level power distribution controller can centrally distribute the given power F Mk_set to the power shafts of a few power units by cutting the axis to reduce the The overall excitation power consumption of the motor vehicle is improved, thereby improving the efficiency of the whole vehicle.
  • the above-mentioned train power distribution method provided by the present invention may further include the step of: maximizing the execution of a given power distributed to the power unit according to the wheel-rail adhesion state of each power unit.
  • the power execution and observation of the power unit can be implemented by the adhesion utilization control module and the traction inverter control module, mainly for maximizing the physical adhesion of the current power unit, and according to the acceleration of the wheelset. , Creep speed state Feedback the wheel-rail state and maximum allowable power of each power unit to the vehicle-level power distribution controller in real time as its decision-making basis.
  • the adhesion utilization control module and the traction inverter control module can be configured at the power unit level power execution and observation control layer (ie, the third control layer).
  • the level of control at this level depends on the smallest control unit of the powered vehicle being controlled. For example, in a frame-controlled vehicle, the level at which this layer of control is located is the bogie unit; in an axle-controlled vehicle, the level at which this layer of control is located is each power axle.
  • the main input signals of the above power unit level power execution and observation control layer include the power command F uj_set issued by the vehicle level power distribution controller to the control unit, and its main output signals are the wheel-rail adhesion state coefficient ⁇ uj , the maximum power exertion capacity F uj_limit , motor comprehensive speed ⁇ uj and motor comprehensive temperature T uj .
  • the above-mentioned adhesion control module can observe the creep speed of the wheelset and the acceleration of the wheelset in real time to calculate the wheel-rail adhesion state coefficient of the power unit:
  • v uj_creep (t) is the creep Slip speed, indicating the difference between the wheelset speed and the train reference speed;
  • a uj_adh (t) is the wheelset acceleration index, indicating the difference between the wheelset acceleration and the train reference acceleration.
  • 0 of ⁇ uj is the critical point of idling
  • ⁇ uj ⁇ 0 indicates that there is a tendency of idling or idling has occurred
  • ⁇ uj > 0 indicates that the creep speed and acceleration are within the normal range There is no idling sliding trend.
  • the sticking control module can determine the wheelset idling sliding state of the corresponding power unit according to the value of the wheel-rail sticking state coefficient ⁇ uj .
  • the sticking control module can judge that the wheelset of the power unit has no idling sliding tendency, and can fully exert the given power value. At this time, the sticking utilization control module can directly use the given power F uj_set distributed to the power unit as the sticking given force F adh , and send it to the back-end inverter controller.
  • the adhesion utilization control module can use the following formula to calculate the wheelset adhesion force observation feedback at the current time t:
  • F uj_set (t-1) is the given power at the previous moment
  • ⁇ uj is the wheel-rail adhesion state coefficient
  • F max is the maximum allowable power at the current speed of the shaft end of the power unit.
  • the sticking control module can determine that the wheelset of the power unit has a tendency of idling or idling has occurred.
  • the adhesion utilization control module can appropriately adjust the adhesion given force F adh sent to the inverter control through the optimal creep control, fuzzy control, phase method control, and sliding mode variable control, etc.
  • the wheel-rail adhesion state of the wheelset controls the vicinity of its optimal adhesion point.
  • FIG. 4 shows a schematic diagram of an optimal adhesion point provided according to some embodiments of the present invention.
  • the train can store multiple curves of relationship between creep rate and adhesion coefficient.
  • the abscissa of this relationship is the creep rate, which indicates the ratio of the creep speed v uj_creep (t) to the train reference speed.
  • the ordinate of the relationship curve is the adhesion coefficient, which indicates the ratio of wheel-rail adhesion to axle weight.
  • Each relationship curve indicates the variation of the adhesion coefficient with the creep rate under a road condition, and the highest point is the optimal adhesion point under the road condition.
  • the adhesion utilization control module can call the corresponding relationship curve according to the current specific road conditions of the motor vehicle to query the optimal creep rate under the road conditions, so as to calculate the corresponding wheelset adhesion force F uj_adh :
  • J is the moment of inertia of the wheel set
  • v uj_w is the wheel set speed
  • r uj is the effective wheel diameter
  • F m is the actual force exerted by the motor.
  • the wheelset adhesion force F uj_adh obtained by calculation is smaller than the given power F uj_set distributed to the power unit.
  • the adhesion control module can send the wheelset adhesion force F uj_adh to the inverter controller as the adhesion given force F adh .
  • the sticking utilization control module can select the maximum value of the wheelset sticking force F uj_adh in a complete idling coasting control cycle, and feed it back to the vehicle-level power distribution controller of the host motor vehicle for composing the vehicle state of the host motor vehicle.
  • the adhesion control module can simultaneously observe the adhesion state and adhesion force of all wheelsets, and take the minimum value as the The adhesion state and adhesion force fed back by this power unit.
  • the power execution and observation control layer of the power unit may also include a traction inverter control module.
  • the main function of the traction inverter control module is to control the actual torque exerted by the traction motor to a given adhesion force F adh , and collect the temperature T uj , current, voltage and rotational speed ⁇ uj of the traction motor in real time to determine the traction of the power unit. Inverted state.
  • the traction inverter control module can determine that there is no need to perform power limitation, so as to control the traction motor to perform sticking and utilize the sticking power issued by the control module. Concentration Fadh .
  • the traction inverter control module needs to limit the power of the traction motor according to the traction inverter state, and calculate the power of the traction motor after the limit is calculated.
  • the traction inverter control module can compare the issued adhesion given force F adh with the calculated limit power F uj_inv , and control the traction motor to execute the smaller value.
  • the traction inverter control module can directly feed back the traction motor temperature Tuj of the power shaft. If the number of control axes of the power unit is greater than 1, the traction inverter control module can simultaneously collect the temperatures of all traction motors, and feed back the average value as the comprehensive motor temperature T uj .
  • the traction inverter control module can directly feed back the traction motor speed ⁇ uj of the power axis. If the number of control axes of the power unit is greater than 1, the traction inverter control module can simultaneously collect the rotational speeds of all traction motors, and feed back the average value as the comprehensive motor rotational speed ⁇ uj .
  • the traction inverter control module can use the following formula to calculate the smaller value of the adhesion given force F adh and the power-limiting power F uj_inv :
  • the traction inverter control module can feed back the smaller value of the sticking given force F adh and the power-limiting power F uj_inv to the vehicle-level power distribution control as the unit maximum allowable power F uj_limit of the present power unit.
  • the above-mentioned train power distribution method provided by the present invention can build a three-layer controller based on the existing train-vehicle-power unit control hierarchy for intelligent collaborative distribution of train power.
  • the present invention can realize multi-objective traction optimization control such as maximum traction performance, minimum longitudinal impact, and optimal converter system state (efficiency, temperature rise) of the automatic driving train. , so as to solve the problems that the existing locomotive power synchronous distribution method is easy to cause limited power exertion and longitudinal impact of the train under severe conditions such as long marshalling and complex curves.
  • controllers described in the above embodiments can be implemented by a combination of software and hardware. It will be appreciated, however, that these controllers may also be implemented solely in software or hardware.
  • these controllers can be implemented in one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), processors,
  • a controller, microcontroller, microprocessor, other electronic device for performing the above-described functions, or a selected combination of the above-described devices is implemented.
  • these controllers may be implemented by separate software modules such as procedures and functions running on a general-purpose chip, each of which may perform one or more of the Describes the functions and operations.

Landscapes

  • Engineering & Computer Science (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Electric Propulsion And Braking For Vehicles (AREA)
  • Current-Collector Devices For Electrically Propelled Vehicles (AREA)

Abstract

公开了一种列车动力分配方法及装置，以及一种计算机可读存储介质。该方法包括：根据列车的行驶规划曲线、行驶线路信息及各动车的车辆状态，以最小纵向冲击为目标，将列车总动力分配到各动车；根据各动车的多个动力单元的状态，以最佳变流器系统状态为目标，将分配到本动车的给定动力进一步分配到本动车的各动力单元；以及根据各动力单元的轮轨粘着状态，最大化地执行分配到本动力单元的给定动力。根据列车各编组的动力学状态对各编组车辆的动力分配进行异步协同优化，从而实现列车的最大牵引发挥、最小纵向冲击、最佳变流器系统状态等多目标的牵引优化控制。

Description

一种列车动力分配方法及装置 技术领域
本发明涉及轮轨式列车的动力分配技术，尤其涉及一种列车动力分配方法，以及一种列车动力分配装置。
背景技术
牵引力与电制动力控制是列车控制的核心功能之一，其控制性能直接影响着列车运行的安全性和可靠性。受限于传统的人工驾驶方式，目前多动力单元列车的动力分配通常采用同步均分的方式，即司机通过控制手柄给定列车总需求动力，再由列车控制系统按照平均分配的方式将列车总需求动力下发为每个动力单元的给定力矩。
然而，列车动力发挥受制于轮轨间物理粘着系数。随着列车牵引吨位、编组长度的不断加大，列车的动力分布也更多地呈现为分散式，而且前序动力单元的轮对自清扫作用会改善后续动力单元的轨面状态。因此，各动力单元所处的轮轨粘着条件会呈现出较大的差异性。在这种条件下，现有的同步均分式动力给定模式不仅会增加粘着条件较差机车的空转频率、降低其牵引发挥平稳性，还会限制粘着条件较好的机车牵引发挥效率。
此外，大编组列车在连续弯道或坡道运行时，列车编组呈多变起伏曲线分布，处于不同坡道、弯道上的动力单元实际所需动力各不相同。例如：为了使列车纵向冲击相对最小(即使各车辆之间的车钩力最小)，可能需要向处于上坡道的动力单元提供牵引力，而向处于下坡道的动力单元提供电制动力。随着列车编组长度与运行速度的不断提升，列车动力分配对列车动力学状态的影响越来越大。在一些极端工况下，采用的动力同步均分方式会导致动力协同不佳，从而引起车钩力剧增，甚至导致车钩断裂的严重事故。
为了克服现有技术的上述缺陷，本领域亟需一种列车动力的分配技术，用于根据列车各编组的动力学状态对各编组车辆的动力分配进行异步协同优化，从而实现列车的最大牵引发挥、最小纵向冲击、最佳变流器系统状态等多目标的牵引优化控制。
发明内容
以下给出一个或多个方面的简要概述以提供对这些方面的基本理解。此概述不是所有构想到的方面的详尽综览，并且既非旨在指认出所有方面的关键性或决定性要素 亦非试图界定任何或所有方面的范围。其唯一的目的是要以简化形式给出一个或多个方面的一些概念以为稍后给出的更加详细的描述之前序。
为了克服现有技术的上述缺陷，本发明提供了一种列车动力分配方法、一种列车动力分配装置，以及一种计算机可读存储介质，用于根据列车各编组的动力学状态对各编组车辆的动力分配进行异步协同优化，从而实现列车的最大牵引发挥、最小纵向冲击、最佳变流器系统状态等多目标的牵引优化控制。
在本发明提供的上述列车动力分配方法中，所述列车包括多节车辆。所述多节车辆分为动车和拖车。所述列车动力分配方法包括：根据所述列车的行驶规划曲线、行驶线路信息及各所述动车的车辆状态，以最小纵向冲击为目标，将列车总动力分配到各所述动车；根据各所述动车的多个动力单元的状态，以最佳变流器系统状态为目标，将分配到本动车的给定动力进一步分配到本动车的各所述动力单元；以及根据各所述动力单元的轮轨粘着状态，最大化地执行分配到本动力单元的给定动力。
优选地，在本发明的一些实施例中，所述行驶规划曲线可以包括规划速度曲线及规划动力曲线，用于指示所述列车在行驶线路的各时刻的列车速度及列车总动力。所述行驶线路信息可以包括所述列车在当前时刻的坡道坡度及曲线半径。各所述动车的车辆状态可以包括各所述动车反馈的动车最大允许动力。将所述列车总动力分配到各所述动车的步骤可以包括：以列车总动力分配方案为求解对象，根据所述行驶规划曲线、所述行驶线路信息及各所述动车的车辆状态，对所述列车的纵向冲击进行量化建模，其中，所述列车总动力分配方案指示分配到各所述动车的动力；以及对构建的纵向冲击量化模型进行最优化求解，以获取最小纵向冲击对应的列车总动力分配方案。
优选地，在本发明的一些实施例中，对所述列车的纵向冲击进行量化建模的步骤可以包括：根据所述行驶规划曲线、所述行驶线路信息及各所述动车的车辆状态，计算各所述车辆之间的车钩力及车钩力冲量；以及根据各所述车辆之间的最大车钩力及最大车钩力冲量量化所述列车的纵向冲击，以构建所述纵向冲击量化模型。
可选地，在本发明的一些实施例中，对所述纵向冲击量化模型进行最优化求解的步骤可以包括：采用控制变量参数化方法或群体智能算法，在各所述动车的单周期动力最大允许变化量的范围内，对所述列车总动力分配方案进行最优化求解，其中，所述动车的单周期动力最大允许变化量由所述动车的车辆速度和/或变流器系统状态决定。
可选地，在本发明的一些实施例中，所述列车动力分配方法还可以包括：先根据 列车的动力单周期最大允许变化量，对所述规划动力曲线的各时刻的列车总动力进行限幅滤波处理，其中，所述列车的动力单周期最大允许变化量由列车速度、列车网压和/或行驶线路条件决定；以及再根据所述限幅滤波处理后的规划动力曲线，对所述列车的纵向冲击进行量化建模。
优选地，在本发明的一些实施例中，所述列车动力分配方法还可以包括：根据各所述动车的动车最大允许动力计算对应时刻的列车最大允许动力；响应于所述列车最大允许动力小于所述限幅滤波处理后的规划动力曲线的对应时刻的列车总动力，将由各所述动车的动车最大允许动力构成的列车动力分配方案代入所述纵向冲击量化模型，以计算对应的列车冲击量化值；响应于所述列车冲击量化值小于量化门槛值，根据各所述动车的动车最大允许动力分配所述列车总动力，其中，所述量化门槛值是根据列车运行安全评估得到的最大允许纵向冲击量化值决定；以及响应于所述列车冲击量化值大于或等于所述量化门槛值，逐步减小所述对应时刻的列车总动力，直到所述列车冲击量化值小于所述量化门槛值。
可选地，在本发明的一些实施例中，所述动车的车辆状态可以包括本动车的各所述动力单元反馈的轮轨粘着状态系数、单元最大允许动力、电机综合转速和/或电机综合温度。将分配到本动车的给定动力进一步分配到本动车的各所述动力单元的步骤可以包括：响应于本动车的动车最大允许动力大于或等于分配到本动车的给定动力，以动车动力分配方案为求解对象，根据本动车的所述车辆状态对变流器系统状态进行量化建模，其中，所述动车动力分配方案指示分配到本动车的各所述动力单元的动力，所述动车最大允许动力是根据各所述动力单元的单元最大允许动力计算；以及在本动车的各所述动力单元的单周期动力最大允许变化量的范围内，对构建的系统状态量化模型进行最优化求解，以获取最佳变流器系统状态对应的动车动力分配方案，其中，所述动力单元的单周期动力最大允许变化量由所述动车的车辆速度和/或变流器系统状态决定。
优选地，在本发明的一些实施例中，将分配到本动车的给定动力进一步分配到本动车的各所述动力单元的步骤还可以包括：响应于本动车的动车最大允许动力小于分配到本动车的给定动力，以动车动力分配方案为求解对象，根据本动车的动力发挥目标函数及变流器状态目标函数，对本动车的动力发挥及变流器系统状态进行综合的量化建模，其中，所述动力发挥目标函数指示分配到本动车的各所述动力单元的动力之和，所述变流器状态目标函数指示变流器系统状态的量化值；在本动车的各所述动力 单元的单周期动力最大允许变化量的范围内，对构建的综合量化模型进行最优化求解，以获取最优综合情况对应的最优集前沿面；以及根据最低牵引力限制准则、最低变流器状态限制准则及牵引力发挥与变流器状态优先级，从所述最优集前沿面的多个动车动力分配方案中选取对应的最优解。
可选地，在本发明的一些实施例中，将分配到本动车的给定动力进一步分配到本动车的各所述动力单元的步骤还可以包括：响应于分配到本动车的给定动力小于动力门槛值，将分配到本动车的给定动力集中分配到本动车的部分动力单元，其中，所述动力门槛值是根据本动车的各所述动力单元的能量效率决定。
可选地，在本发明的一些实施例中，最大化地执行分配到本动力单元的给定动力的步骤可以包括：根据轮对的蠕滑速度及轮对加速度指标，计算本动力单元的轮轨粘着状态系数；响应于所述轮轨粘着状态系数指示所述轮对无空转滑行趋势，将分配到本动力单元的给定动力作为粘着给定力下发到逆变控制器；响应于所述轮轨粘着状态系数指示所述轮对有空转滑行趋势或已发生空转滑行，根据所述动力单元的有效轮径、轮对转动惯量、轮对速度及电机实际发挥力计算轮对粘着力，并将所述轮对粘着力作为粘着给定力下发到所述逆变控制器；以及以所述逆变控制器控制本动力单元的牵引电机执行所述粘着给定力。
优选地，在本发明的一些实施例中，所述列车动力分配方法还可以包括：将本动力单元的所述轮轨粘着状态系数反馈到对应动车的车辆级控制器，以用于组成所述对应动车的车辆状态；响应于所述轮轨粘着状态系数指示所述轮对无空转滑行趋势，根据本动力单元轴端当前转速下的最大允许动力、前一时刻的给定动力及所述轮轨粘着状态系数计算当前时刻的轮对粘着力，并将所述当前时刻的轮对粘着力反馈到所述车辆级控制器，以用于组成所述对应动车的车辆状态；以及响应于所述轮轨粘着状态系数指示所述轮对有空转滑行趋势或已发生空转滑行，选取一个完整空转滑行控制周期内所述轮对粘着力的最大值反馈到所述车辆级控制器，以用于组成所述对应动车的车辆状态。
可选地，在本发明的一些实施例中，最大化地执行分配到本动力单元的给定动力的步骤还可以包括：采集本动力单元的牵引电机的温度、电流、电压及转速，以判定本动力单元的牵引逆变状态；响应于所述牵引逆变状态良好，控制所述牵引电机执行所述粘着给定力；以及响应于所述牵引逆变状态不佳，根据所述牵引逆变状态对所述牵引电机的功率进行限制并计算对应的限功动力，对所述粘着给定力与所述限功动力 进行比较，以控制牵引电机执行其中的较小值。
优选地，在本发明的一些实施例中，所述动力单元可以包括至少一根控制轴。每根控制轴对应至少一台牵引电机。所述动力分配方法还可以包括：对各所述牵引电机的转速取平均值以作为所述动力单元的电机综合转速，将所述电机综合转速反馈到对应动车的车辆级控制器，以用于组成所述对应动车的车辆状态；对各所述牵引电机的温度取平均值以作为所述动力单元的电机综合温度，将所述电机综合温度反馈到所述车辆级控制器，以用于组成所述对应动车的车辆状态；以及将所述粘着给定力与所述限功动力中的较小值作为所述动力单元的单元最大允许动力，并将所述单元最大允许动力反馈到所述车辆级控制器，以用于组成所述对应动车的车辆状态。
可选地，在本发明的一些实施例中，所述列车动力分配方法还可以包括：从所述列车的自动驾驶系统获取所述行驶规划曲线；从列车运行监控记录装置获取所述行驶线路信息；以及从各所述动车的车辆级控制器获取各所述动车的车辆状态，其中，各所述车辆状态分别由对应动车的多个动力单元反馈到所述车辆级控制器的单元状态组合而成。
根据本发明的另一方面，本文还提供了一种列车动力分配装置。
在本发明提供的上述列车动力分配装置中，所述列车包括多节车辆。所述多节车辆分为动车和拖车。所述列车动力分配装置包括存储器及处理器。所述处理器连接所述存储器，并配置用于实施上述任意一个实施例所提供的列车动力分配方法。
根据本发明的另一方面，本文还提供了一种计算机可读存储介质。
本发明提供的上述计算机可读存储介质，其上存储有计算机指令。所述计算机指令被处理器执行时，可以实施上述任意一个实施例所提供的列车动力分配方法。
附图说明
在结合以下附图阅读本公开的实施例的详细描述之后，能够更好地理解本发明的上述特征和优点。在附图中，各组件不一定是按比例绘制，并且具有类似的相关特性或特征的组件可能具有相同或相近的附图标记。
图1示出了根据本发明的一些实施例提供的列车动力分配方法的流程示意图。
图2示出了根据本发明的一些实施例提供的列车纵向受力的示意图。
图3示出了根据本发明的一些实施例提供的最优集前沿面的示意图。
图4示出了根据本发明的一些实施例提供的最佳粘着点的示意图。
具体实施方式
以下由特定的具体实施例说明本发明的实施方式，本领域技术人员可由本说明书所揭示的内容轻易地了解本发明的其他优点及功效。虽然本发明的描述将结合优选实施例一起介绍，但这并不代表此发明的特征仅限于该实施方式。恰恰相反，结合实施方式作发明介绍的目的是为了覆盖基于本发明的权利要求而有可能延伸出的其它选择或改造。为了提供对本发明的深度了解，以下描述中将包含许多具体的细节。本发明也可以不使用这些细节实施。此外，为了避免混乱或模糊本发明的重点，有些具体细节将在描述中被省略。
如上所述，现有的同步均分式动力给定模式不仅会增加粘着条件较差机车的空转频率、降低其牵引发挥平稳性，还会限制粘着条件较好的机车牵引发挥效率。为了克服现有技术的上述缺陷，本发明提供了一种列车动力分配方法、一种列车动力分配装置，以及一种计算机可读存储介质。本发明能够基于列车各动力单元的动力学状态、轮轨状态及变流器状态等综合约束，进行列车动力(此处动力含牵引力、电制动力，后文均采用动力表称牵引力、电制动力)的异步协同优化，从而实现自动驾驶列车的最大牵引发挥、最小纵向冲击、最佳变流器系统状态等多目标牵引优化控制。
在一些非限制性的实施例中，上述列车动力分配方法可以由一种列车动力分配装置来自动实施。具体来说，该列车动力分配装置可以包括存储器及处理器。该存储器可以包括计算机可读存储介质，其上存储有计算机指令。该处理器可以连接该存储器，并配置用于执行该存储器上存储的计算机指令以实施上述列车动力分配方法。
以下将结合一些利用上述列车动力分配装置实施上述列车动力分配方法的实施例来描述本发明的主要发明构思。本领域的技术人员可以理解，这些列车动力分配方法的实施方式只是一些非限制性的实施例，旨在清楚地展示本发明的主要发明构思，并提供一些便于公众实施的具体方案，而非用于限制本发明的保护范围。
请参考图1，图1示出了根据本发明的一些实施例提供的列车动力分配方法的流程示意图。
如图1所示，本发明提供的上述列车动力分配方法可以包括步骤：根据列车的行驶规划曲线、行驶线路信息及各动车的车辆状态，执行列车级动力协同控制。
在本发明的一些实施例中，上述列车级动力协同控制可以由列车级动力协同控制器实施。该列车级动力协同控制器可以由列车动力分配装置的处理器实现，与列车的 自动驾驶系统主机同处于列车的第一控制层次，主要用于根据自动驾驶系统输入的规划曲线、列车运行监控记录装置(LKJ)输入的线路信息及各动车反馈的车辆状态，预测列车的纵向动力学状态，从而以最小纵向冲击为目标，将列车总动力最优化的分配到各动车。
具体来说，列车级动力协同控制器可以首先从列车的自动驾驶系统获取其规划的行驶规划曲线，从列车运行监控记录装置(LKJ)获取当前时刻的行驶线路信息，并从各动车的车辆级控制器获取各动车的车辆状态，以作为列车级动力协同控制的输入信息。在一些实施例中，行驶规划曲线可以包括规划速度曲线V train＝{v train(t)v train(t+1)…v train(t+n)}及规划动力曲线F train＝{F train(t)F train(t+1)…F train(t+n)}。该规划速度曲线V train可以包括多个元素，每个元素指示列车在行驶线路的对应时刻的列车速度。该规划动力曲线F train可以包括多个元素，每个元素指示列车在行驶线路的对应时刻的列车总动力。在一些实施例中，行驶线路信息可以包括列车在当前时刻的坡道坡度i、坡道长度l i、曲线半径R及曲线长度l r。在一些实施例中，各动车的车辆状态可以包括各动车的车辆级控制器反馈的动车最大允许动力F limit＝{F M1_limit…F Mn_limit…F Mlast_limit}。该动车最大允许动力F limit包括多个元素，每个元素指示一节动车的动车最大允许动力(例如：F M1_limit)。该一节动车的动车最大允许动力F M1_limit，可以通过对动车M1的多个动力单元反馈到其车辆级控制器的多个单元最大允许动力F uj_limit之和来计算。
之后，列车级动力协同控制器可以根据上述输入信息执行列车级动力协同控制，将列车总动力异步协同地分配到列车的各动车，以解决现有机车动力同步分布方法在大编组、复杂曲线等恶劣工况下易造成动力发挥受限的问题。
在一些优选的实施例中，为防止列车的纵向冲击及电网冲击过大，列车级动力协同控制器可以首先对接收的规划动力曲线F train进行限幅滤波处理：
Figure PCTCN2020130343-appb-000001
式中：
Figure PCTCN2020130343-appb-000002
为滤波处理后t时刻的列车规划动力；f dec为列车动力单周期最大允许下降量，由系统控制列车性能决定，可为固定值亦可为与列车速度、列车网压、线路条件等实时状态有关的函数输出值；f ris为列车动力单周期最大允许上升量，由系统控制列车性能决定，可为固定值亦可为与列车速度、列车网压、线路条件等实时状态 有关的函数输出值。通过设置f dec和f ris来限制整车的牵引力变化及牵引逆变单元的功率变化，可以有效地降低每个控制周期之间的牵引力变化量及牵引功率变化量，从而起到降低列车的纵向冲击及电网冲击的效果。
请参考图2，图2示出了根据本发明的一些实施例提供的列车纵向受力的示意图。
如图2所示，在本发明的一些实施例中，列车可以包括多节车辆，分为动车和拖车。M为动车标识，T为拖车标识，动车和拖车可以采用连续编号形式进行编号(即M1、T2……Tn-1、Mn、Tn+1……)。
在一些实施例中，列车级动力协同控制器可以将列车总动力分配方案F set＝{F M1_set…F Mn_set…F Mlast_set}作为求解对象，根据限幅滤波处理后的规划动力曲线
Figure PCTCN2020130343-appb-000003
规划速度曲线V train、列车在当前时刻的坡道坡度i、坡道长度l i、曲线半径R、曲线长度l r及各动车的动车最大允许动力F limit，对列车的纵向冲击进行量化建模：
Figure PCTCN2020130343-appb-000004
在上述公式(2)指示的纵向冲击量化模型中，
Figure PCTCN2020130343-appb-000005
为各车辆之间的最大车钩力；
Figure PCTCN2020130343-appb-000006
为各车辆之间的最大车钩力冲量；K P,K I分别是最大车钩力与最大车钩力冲量的可调增益。
Figure PCTCN2020130343-appb-000007
指示列车纵向冲击的量化值，量化值越小，列车的纵向冲击也越小。函数min J[F set(t)]用于计算能够使
Figure PCTCN2020130343-appb-000008
取得最小值的列车总动力分配方案F set。该列车总动力分配方案F set包括多个元素，每个元素指示分配到一节动车的动力。
进一步地，Mk,Tk中得M为动车标识；T为拖车标识；k为车辆编号；Mlast为动车的最大编号。
Figure PCTCN2020130343-appb-000009
为车钩的钩缓模型，其输入量为车钩力的历史序列，而其输出量为车钩增量的预测值。f k(t)＝f i+f r+M k(a k+b kv+c kv 2)为车辆阻力，其中， f i＝M ki k为坡道阻力；f r＝M k(600/R k)为曲线阻力；i k为k车所处线路的坡道坡度；R k为k车所处线路曲线半径。在一些实施例中，i k和R k可以根据列车在当前时刻的坡道坡度i、坡道长度l i、曲线半径R、曲线长度l r及每节车辆的长度计算获得。M k(a+bv+cv 2)为k车的运行阻力，其中，M k为k车质量；动力车辆的阻力系数(a M,b M,c M)由离线辨识得到；拖车采用平均拖车阻力系数(a T,b T,c T)由递推最小二乘法在线辨识得到。a Mk(t)为动力车辆在t时刻的瞬时加速度，通过轮轴速度传感器信号经数值处理得到。a Tk(t)为拖车参考加速度。在一些实施例中，a Tk(t)可以通过拖车的速度传感器信号经数值处理得到。可选地，在另一些实施例中，a Tk(t)可以采用列车参考加速度，由列车参考速度经数值微分处理得到。
在一些实施例中，拖车的平均拖车阻力系数(a T,b T,c T)可以根据如下稳态模型，通过递推最小二乘法在线参数辨识得到：
Figure PCTCN2020130343-appb-000010
式中：M Tall为总拖车质量，即列车的牵引载荷。在使用递推最小二乘在线参数辨识时，可以选取：
Figure PCTCN2020130343-appb-000011
θ＝[a T,b T,c T]                      (5)
Figure PCTCN2020130343-appb-000012
式中：θ为递推最小二乘在线参数辨识的辨识参数；Φ为列车的当前状态量。
本领域的技术人员可以理解，上述基于最大车钩力
Figure PCTCN2020130343-appb-000013
及最大车钩力冲量
Figure PCTCN2020130343-appb-000014
来构建纵向冲击量化模型的方案，只是本发明提供的一种非限制性的案例，旨在清楚地展示本发明的主要构思，并提供一种便于公众实施的具体方案，而非用于限制本发明的保护范围。可选地，在另一些实施例中，本领域的技术人员也可以基于本发明的构思，采用速度、加速度以及加速度微分等指标来替代上述最大车钩力
Figure PCTCN2020130343-appb-000015
及最大车钩力冲量
Figure PCTCN2020130343-appb-000016
以构建对应的纵向冲击量化模型，从而同样达到量化列车纵向冲击的效果。
在建立纵向冲击量化模型之后，列车级动力协同控制器可以采用控制变量参数化方法(Control Variable Parameterization，CVP)或群体智能算法(例如：PSO粒子群 算法、TLBO教与学算法)，在列车总动力分配方案F set的可行解域中对构建的纵向冲击量化模型进行最优化求解，以获取最小纵向冲击对应的列车总动力分配方案F set。列车总动力分配方案F set的可行解域为F k_set(t)≤F k_limit(t)&F k_set(t-1)-f vk_dec≤F k_set(t)≤F k_set(t-1)+f vk_ris，其中，k为动力车辆编号；f vk_dec为动力车辆k的单周期动力最大允许下降量，由车辆牵引变流性能和列车冲击率限制决定，可为固定值亦可为与车辆速度、变流器状态等实时状态有关的函数输出值；f vk_ris为动力车辆k的单周期动力最大允许上升量，由车辆牵引变流性能和列车冲击率限制决定，可为固定值亦可为与车辆速度、变流器状态等实时状态有关的函数输出值。
本领域的技术人员可以理解，上述控制变量参数化方法及群体智能算法只是本发明提供的一种非限制性的案例，旨在清楚地展示本发明的主要构思，并提供一种便于公众实施的具体方案，而非用于限制本发明的保护范围。可选地，在另一些实施例中，本领域的技术人员也可以基于本发明的构思，采用神经网络、深度学习等其他算法对纵向冲击量化模型进行最优化求解，以同样计算获得最小纵向冲击对应的列车总动力分配方案F set。
优选地，在本发明的一些实施例中，列车级动力协同控制器可以进一步根据各动车的动车最大允许动力计算对应时刻的列车最大允许动力，并将该列车最大允许动力与限幅滤波处理后的规划动力曲线的对应时刻的列车总动力
Figure PCTCN2020130343-appb-000017
进行对比，以判断两者的大小。若
Figure PCTCN2020130343-appb-000018
则列车级动力协同控制器可以判断自动驾驶系统给定的列车总动力
Figure PCTCN2020130343-appb-000019
未超过列车最大允许动力的限制，上述列车总动力分配方案F set可以实现。
反之，若
Figure PCTCN2020130343-appb-000020
则列车级动力协同控制器可以判断自动驾驶系统给定的列车总动力
Figure PCTCN2020130343-appb-000021
超过列车最大允许动力的限制，上述列车总动力分配方案F set无法实现。此时，列车级动力协同控制器需要将由各动车的动车最大允许动力构成的列车动力分配方案F limit代入上述纵向冲击量化模型，以计算对应的列车冲击量化值J[F limit(t)]。之后，列车级动力协同控制器可以根据该列车冲击量化值J[F limit(t)]与预设的量化门槛值J limit的大小。该量化门槛值J limit可以根据列车运行安全评估得到的最大允许纵向冲击量化值决定，主要用于保障列车的安全运行。若J[F limit(t)]<J limit， 则说明列车动力分配方案F limit符合列车运行安全评估的要求，列车级动力协同控制器可以根据各动车的动车最大允许动力F limit分配列车总动力。
反之，若J[F limit(t)]≥J limit，则说明列车动力分配方案F limit不符合列车运行安全评估的要求。列车级动力协同控制器需要按预设的固定增量为步长逐步减少
Figure PCTCN2020130343-appb-000022
直到
Figure PCTCN2020130343-appb-000023
此时，列车级动力协同控制器可以求得可行解，使得列车冲击量化值小于量化门槛值，即J[F limit(t)]<J limit。在一些实施例中，列车级动力协同控制器可以将减少后的
Figure PCTCN2020130343-appb-000024
反馈至列车的自动驾驶系统，从而由自动驾驶系统将根据反馈的
Figure PCTCN2020130343-appb-000025
实时修正动力设定曲线
Figure PCTCN2020130343-appb-000026
如图1所示，本发明提供的上述列车动力分配方法还可以包括步骤：根据各动车的多个动力单元的状态，执行车辆级动力分配控制。
在本发明的一些实施例中，上述车辆级动力分配控制可以由车辆级动力分配控制器实施。该车辆级动力分配控制器可以由列车动力分配装置的处理器实现，配置于各动车的车辆网络或逻辑控制层次(即第二控制层次)。该车辆级动力分配控制器主要用于根据各动力单元的轮轨粘着状态、各动力牵引变流系统的效率与温升，以及各动力单元健康状态，以最佳变流器系统状态为目标，将列车级动力协同层分配的给定动力最优化的分配到本动车各动力单元。
具体来说，车辆级动力分配控制器可以首先获取本动车的各动力单元反馈的动力单元状态，以组成本动车的车辆状态。在一些实施例中，动车的车辆状态可以包括本动车的各动力单元的轮轨粘着状态系数α＝[α u1,...,α uj,...,α uN]、单元最大允许动力F u_limit＝[F u1_limit,...,F uj_limit,...,F uN_limit]、电机综合转速ω＝[ω u1,...,ω uj,...,ω uN]和/或电机综合温度T＝[T u1,...,T uj,...,T uN]。该轮轨粘着状态系数α包括多个元素，每个元素指示本动车的一个动力单元的轮轨粘着状态系数。该单元最大允许动力F u_limit包括多个元素，每个元素指示本动车的一个动力单元的单元最大允许动力。该电机综合转速ω包括多个元素，每个元素指示本动车的一个动力单元的电机综合转速。该电机综合温度T包括多个元素，每个元素指示本动车的一个动力单元的电机综合温度。在一些实施例中，车辆级动力分配控制器可以对单元最大允许动力F u_limit中的各元素求和，以计算本动车的动车最大允许动力F Mk_limit，即：
Figure PCTCN2020130343-appb-000027
式中：N为本动力车辆的动力单元个数，F uj为动力单元j反馈的单元最大允许动力。
之后，车辆级动力分配控制器可以根据本动车的上述车辆状态，以及列车级动力协同层分配到本动车的给定动力F Mk_set，制定本动车的动车动力分配方案F u_set＝{F u1_set…F uk_set…F uN_set}。该动车动力分配方案F u_set包括多个元素，每个元素指示分配到本动车的一个动力单元的给定动力。
在一些实施例中，车辆级动力分配控制器可以根据给定动力F Mk_set与动车最大允许动力F Mk_limit的大小关系，选择合适的分配方案来进行车辆级动力分配控制。
具体来说，若本动车的动车最大允许动力大于或等于分配到本动车的给定动力(即F Mk_limit≥F Mk_set)，车辆级动力分配控制器可以判断给定动力F Mk_set并未超过动车最大允许动力F Mk_limit的限制，本动力车辆可以完全发挥出列车级动力协同控制器下发的给定动力F Mk_set。此时，车辆级动力分配控制器无需考虑动力的最大发挥，仅以变流器最佳状态为目标，根据本动车的车辆状态对变流器系统状态进行量化建模：
Figure PCTCN2020130343-appb-000028
在上述公式(8)指示的系统状态量化模型中，η tol为车辆总效率；K eff为效率指标可调权重增益；ΔT tol为系统总温升；K T为温升指标可调权重增益；S tol为系统噪声；K SIL为系统噪声可调权重增益；β adh为车辆粘着状态，指标越小表示越容易发生空转滑行；K adh为车辆粘着状态可调权重增益。K effη tol+K T/ΔT tol+K SIL/S tol+K adhβ adh为变流器系统状态的量化值，量化值越大，变流器的系统状态就越好。函数max J[F u_set(t)]用 于计算能够使K effη tol+K T/ΔT tol+K SIL/S tol+K adhβ adh取得最大值的动车动力分配方案F u_set。
进一步地，η uj为动力单元j的预测效率；r uj为动力单元j的有效轮径。Model eff为动力单元效率预测模型，为机理模型或基于试验数据的经验模型。Model Tep为动力单元的温升预测模型，为机理模型或基于试验数据的经验模型。Model SIL为动力单元的噪声预测模型，为机理模型或基于试验数据的经验模型。
本领域的技术人员可以理解，上述基于轮轨粘着状态系数α、单元最大允许动力F u_limit、电机综合转速ω和电机综合温度T来构建系统状态量化模型的方案，只是本发明提供的一种非限制性的案例，旨在清楚地展示本发明的主要构思，并提供一种便于公众实施的具体方案，而非用于限制本发明的保护范围。可选地，在另一些实施例中，本领域的技术人员也可以基于本发明的构思，采用变流器寿命指标、轮对磨耗指标等变流器状态指标来量化变流器的系统状态，以构建对应的系统状态量化模型。
在建立系统状态量化模型之后，车辆级动力分配控制器可以采用控制变量参数化方法(Control Variable Parameterization，CVP)或群体智能算法(例如：PSO粒子群算法、TLBO教与学算法)，在动车动力分配方案F u_set的可行解域中对构建的系统状态量化模型进行最优化求解，以获取最佳变流器系统状态对应的动车动力分配方案F u_set。动车动力分配方案F u_set的可行解域为F uj_set(t)≤F uj_limit(t)&F uj_set(t-1)-f uj_dec≤F uj_set(t)≤F uj_set(t-1)+f uj_ris，其中，f uj_dec为动力单元j的单周期下降量限幅，由动力单元性能以及车辆冲击率限制决定，可为固定值亦可为与车辆速度、变流器状态等实时状态有关的函数输出值；f uj_ris为动力单元j的单周期上升量限幅，由车辆牵引变流性能决定，可为固定值亦可为与车辆速度、变流器状态等实时状态有关的函数输出值。
可选地，在另一些实施例中，本领域的技术人员也可以基于本发明的构思，采用神经网络、深度学习等其他算法对上述系统状态量化模型进行最优化求解，以同样计算获得最佳变流器系统状态对应的动车动力分配方案F u_set。
反之，若本动车的动车最大允许动力小于分配到本动车的给定动力(即F Mk_limit＜F Mk_set)，车辆级动力分配控制器可以判断给定动力F Mk_set超过动车最大允许动力F Mk_limit的限制，本动车无法满足列车级动力协同控制器下发的给定动力F Mk_set。此时，车辆级动力分配控制器需要同时考虑最大动力发挥与变流器最佳状态两个优化 目标，根据本动车的动力发挥目标函数及变流器状态目标函数，对本动车的动力发挥及变流器系统状态进行综合的量化建模：
Figure PCTCN2020130343-appb-000029
在上述公式(9)指示的综合量化模型中：f 1(t)为动力发挥目标函数，指示分配到本动车的各动力单元的给定动力之和。f 2(t)为变流器状态目标函数，指示变流器系统状态的量化值。[f 1(t),f 2(t)]指示动力发挥与变流器状态的综合量化值，量化值越大，综合状态就越佳。函数max f(t)用于计算能够使[f 1(t),f 2(t)]取得最大值的动车动力分配方案F u_set。
在建立综合量化模型之后，车辆级动力分配控制器可以采用群体智能算法(例如：PSO粒子群算法、TLBO教与学算法)，在动车动力分配方案F u_set的可行解域中对构建的系统状态量化模型进行最优化求解，以获取最佳综合状态对应的Pareto最优集前沿面。如上所述，动车动力分配方案F u_set的可行解域为F uj_set(t)≤F uj_limit(t)&F uj_set(t-1)-f uj_dec≤F uj_set(t)≤F uj_set(t-1)+f uj_ris。
请参考图3，图3示出了根据本发明的一些实施例提供的最优集前沿面的示意图。
如图3所示，最佳综合状态对应的Pareto最优集前沿面可以包括多个数据点。每个数据点可以指示最佳综合状态对应的一个动车动力分配方案F u_set。车辆级动力分配控制器可以预设的根据最低牵引力限制准则、最低变流器状态限制准则及牵引力发挥与变流器状态优先级，进一步从Pareto最优集前沿面的多个动车动力分配方案中选取一个满足条件的最优解F u_set，以实施车辆级动力分配控制。
可选地，在另一些实施例中，本领域的技术人员也可以基于本发明的构思，采用神经网络、深度学习等其他算法对上述综合量化模型进行最优化求解，以同样计算获得最佳综合状态对应的Pareto最优集前沿面，进而选取一个满足条件的最优解F u_set，以实施上述车辆级动力分配控制。
在一些优选的实施例中，车辆级动力分配控制器可以进一步根据分配到本动车的给定动力F Mk_set及预设的动力门槛值F Mk_th，判断是否需要进行切轴控制以提高整车效率。具体来说，动力门槛值F Mk_th可以由本动车的各动力单元的能量效率决定，用于指示能够使本动车的各动力单元高效运行的最低动力之和。若F Mk_set<F Mk_th，则说明分配到本动车的给定动力较小，车辆级动力分配控制器可以通过切轴控制，将给定动力F Mk_set集中分配到少数动力单元的动力轴，以减少动车整体励磁功率消耗，从而提高整车效率。
如图1所示，本发明提供的上述列车动力分配方法还可以包括步骤：根据各动力单元的轮轨粘着状态，最大化地执行分配到本动力单元的给定动力。
在本发明的一些实施例中，动力单元的动力执行及观测可以由粘着利用控制模块和牵引逆变控制模块配合实施，主要用于最大化发挥当前动力单元的物理粘着，并根据轮对的加速度、蠕滑速度状态将各动力单元的轮轨状态及最大允许动力，实时反馈至车辆级动力分配控制器以作为其决策依据。该粘着利用控制模块和牵引逆变控制模块可以配置于动力单元级动力执行及观测控制层(即第三控制层次)。该层控制所处层次取决于所控制动力车辆的最小控制单元。例如：在架控车辆中，该层控制所处的层次为转向架单元；在轴控车辆中，该层控制所处的层次为各动力轴。
上述动力单元级动力执行及观测控制层的主要输入信号包括车辆级动力分配控制器下发到本控制单元的动力指令F uj_set，而其主要输出信号为轮轨粘着状态系数α uj、最大动力发挥能力F uj_limit、电机综合转速ω uj及电机综合温度T uj。如上所述，车辆级动力分配控制器可以根据本动车的各动力单元反馈的动力单元状态α uj、F uj_limit、ω uj、T uj，组成本动车的车辆状态α＝[α u1,...,α uj,...,α uN]、F u_limit＝[F u1_limit,...,F uj_limit,...,F uN_limit]、ω＝[ω u1,...,ω uj,...,ω uN]、T＝[T u1,...,T uj,...,T uN]。
具体来说，上述粘着利用控制模块可以实时观测轮对的蠕滑速度及轮对加速度情况，以计算本动力单元的轮轨粘着状态系数：
α uj＝Adh_Judge(v uj_creep(t),a uj_adh(t))                 (10)
式中：α uj为轮轨粘着状态为归一化系数，取值范围为[-1 1]；Adh_Judge()为粘着状态判断函数，可以通过模糊规则表来实现；v uj_creep(t)为蠕滑速度，指示轮对速度与列车参考速度的差值；a uj_adh(t)为轮对加速度指标，指示轮对加速度与列车参考加速度的差值。
本领域的技术人员可以理解，上述通过模糊规则表来实现的粘着状态判断函数Adh_Judge()只是本发明提供的一种非限制性的案例，旨在清楚地展示本发明的主要构思，并提供一种便于公众实施的具体方案，而非用于限制本发明的保护范围。可选地，在另一些实施例中，本领域的技术人员也可以基于本发明的构思，使用机理/经验公式、专家规则系统、状态机等策略来实现该粘着状态判断函数Adh_Judge()的计算功能。
在上述公式(10)中，α uj的0值为空转滑行的临界点，α uj＜0表示有空转滑行趋势或已发生空转滑行，α uj＞0表示蠕滑速度和加速度均在正常范围内暂未出现空转滑行趋势。粘着利用控制模块可以根据轮轨粘着状态系数α uj的取值，判定对应动力单元的轮对空转滑行状态。
当α uj＞0时，粘着利用控制模块可以判断本动力单元的轮对无空转滑行趋势，可以完全发挥出给定动力值。此时，粘着利用控制模块可以直接将分配到本动力单元的给定动力F uj_set作为粘着给定力F adh，并下发到后端的逆变控制器。
在一些实施例中，粘着利用控制模块可以采用下式计算当前时刻t的轮对粘着力观测反馈：
F uj_adh(t)＝F uj_set(t-1)+α uj(F max-F uj_set(t-1))           (12)
式中：F uj_set(t-1)为前一时刻的给定动力；α uj为轮轨粘着状态系数；F max为动力单元轴端当前转速下的最大允许动力。之后，粘着利用控制模块可以将计算获得的轮对粘着力F uj_adh反馈到本动车的车辆级控制器，以用于组成本动车的车辆状态。
反之，当α uj＜0时，粘着利用控制模块可以判断本动力单元的轮对有空转滑行趋势或已发生空转滑行。此时，粘着利用控制模块可以通过最优蠕滑控制、模糊控制、相位法控制以及滑模变控制等粘着优化控制策略，适当调整下发给逆变控制的粘着给定力F adh，以将轮对的轮轨粘着状态控制其最佳粘着点附近。
请参考图4，图4示出了根据本发明的一些实施例提供的最佳粘着点的示意图。
如图4所示，列车可以存储有多条蠕滑率与粘着系数的关系曲线。该关系曲线的横坐标为蠕滑率，指示蠕滑速度v uj_creep(t)与列车参考速度的比值。该关系曲线的纵坐标为粘着系数，指示轮轨粘着力与轴重的比值。每条关系曲线指示一种路况下粘着系数随蠕滑率的变化情况，其最高点即为该路况下的最佳粘着点。粘着利用控制模块可以根据动车当前的具体路况，调用对应的关系曲线以查询该路况下的最佳蠕滑率，从而计算对应的轮对粘着力F uj_adh：
Figure PCTCN2020130343-appb-000030
式中：J为轮对转动惯量；v uj_w为轮对速度；r uj为有效轮径；F m为电机实际发挥力。一般情况下，计算获得的轮对粘着力F uj_adh小于分配到本动力单元的给定动力F uj_set。粘着利用控制模块可以将轮对粘着力F uj_adh作为粘着给定力F adh下发到逆变控制器。
在一些实施例中，粘着利用控制模块可以选取一个完整空转滑行控制周期内轮对粘着力F uj_adh的最大值，反馈到本动车的车辆级动力分配控制器以用于组成本动车的车辆状态。
在一些实施例中，若控制车辆的最小动力单元包括多根控制轴(即控制轴数大于1)，粘着利用控制模块可以同时观测所有轮对的粘着状态和粘着力，并取其中最小值作为本动力单元反馈的粘着状态与粘着力。
如上所述，动力单元的动力执行及观测控制层还可以包括牵引逆变控制模块。该牵引逆变控制模块的主要功能是将牵引电机实际发挥力矩控制到粘着给定力F adh，并实时采集牵引电机的温度T uj、电流、电压与转速ω uj，以判定本动力单元的牵引逆变状态。
若牵引电机的温度T uj、电流、电压与转速ω uj指示当前牵引逆变状态良好，则牵引逆变控制模块可以判断无需进行功率限制，从而控制牵引电机执行粘着利用控制模块下发的粘着给定力F adh。反之，若牵引电机的温度T uj、电流、电压与转速ω uj指示当前牵引逆变状态不佳，则牵引逆变控制模块需要根据牵引逆变状态对牵引电机的功率进行限制，并计算限制后的功率对应的限功动力F uj_inv。之后，牵引逆变控制模块可以对下发的粘着给定力F adh与计算获得的限功动力F uj_inv进行比较，控制牵引电机执行其中的较小值。
对于本动力单元反馈给车辆级动力分配控制器的综合电机温度T uj，若本动力单元为单轴控制，则牵引逆变控制模块可以直接反馈该动力轴的牵引电机温度T uj。若动力单元的控制轴数大于1，则牵引逆变控制模块可以同时采集所有牵引电机的温度，并反馈其平均值以作为综合电机温度T uj。
对于本动力单元反馈给车辆级动力分配控制器的综合电机转速ω uj，若本动力单元为单轴控制，则牵引逆变控制模块可以直接反馈该动力轴的牵引电机转速ω uj。若动力单元的控制轴数大于1，则牵引逆变控制模块可以同时采集所有牵引电机的转速，并反馈其平均值以作为综合电机转速ω uj。
对于本动力单元反馈给车辆级动力分配控制器的单元最大允许动力F uj_limit，牵引逆变控制模块可以采用如下公式计算粘着给定力F adh与限功动力F uj_inv中的较小值：
F uj_limit＝min(F uj_inv,F adh)                 (13)
牵引逆变控制模块可以将粘着给定力F adh与限功动力F uj_inv中的较小值反馈到车辆级动力分配控制，以作为本动力单元的单元最大允许动力F uj_limit。
综上所述，本发明提供的上述列车动力分配方法可以基于现有列车-车辆-动力单元的控制层次，构建三层控制器进行列车动力智能协同分配。通过车辆之间、动力单元之间的差异化智能分配，本发明可以实现自动驾驶列车的最大牵引发挥、最小纵向冲击、最佳变流器系统状态(效率、温升)等多目标牵引优化控制，从而解决现有机车动力同步分布方法在长大编组、复杂曲线等恶劣工况下易造成动力发挥受限、列车纵向冲击等问题。
尽管为使解释简单化将上述方法图示并描述为一系列动作，但是应理解并领会，这些方法不受动作的次序所限，因为根据一个或多个实施例，一些动作可按不同次序发生和/或与来自本文中图示和描述或本文中未图示和描述但本领域技术人员可以理解的其他动作并发地发生。
本领域技术人员将可理解，信息、信号和数据可使用各种不同技术和技艺中的任何技术和技艺来表示。例如，以上描述通篇引述的数据、指令、命令、信息、信号、位(比特)、码元、和码片可由电压、电流、电磁波、磁场或磁粒子、光场或光学粒子、或其任何组合来表示。
本领域技术人员将进一步领会，结合本文中所公开的实施例来描述的各种解说性逻辑板块、模块、电路、和算法步骤可实现为电子硬件、计算机软件、或这两者的组 合。为清楚地解说硬件与软件的这一可互换性，各种解说性组件、框、模块、电路、和步骤在上面是以其功能性的形式作一般化描述的。此类功能性是被实现为硬件还是软件取决于具体应用和施加于整体系统的设计约束。技术人员对于每种特定应用可用不同的方式来实现所描述的功能性，但这样的实现决策不应被解读成导致脱离了本发明的范围。
尽管上述的实施例所述的控制器可以通过软件与硬件的组合来实现的。但是可以理解，这些控制器也可以单独在软件或硬件中加以实施。对于硬件实施而言，这些控制器可在一个或多个专用集成电路(ASIC)、数字信号处理器(DSP)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、处理器、控制器、微控制器、微处理器、用于执行上述功能的其它电子装置或上述装置的选择组合来加以实施。对软件实施而言，这些控制器可以通过在通用芯片上运行的诸如程序模块(procedures)和函数模块(functions)等独立的软件模块来加以实施，其中每一个模块可以执行一个或多个本文中描述的功能和操作。
提供对本公开的先前描述是为使得本领域任何技术人员皆能够制作或使用本公开。对本公开的各种修改对本领域技术人员来说都将是显而易见的，且本文中所定义的普适原理可被应用到其他变体而不会脱离本公开的精神或范围。由此，本公开并非旨在被限定于本文中所描述的示例和设计，而是应被授予与本文中所公开的原理和新颖性特征相一致的最广范围。

Claims (16)

  1. 一种列车动力分配方法，其特征在于，所述列车包括多节车辆，所述多节车辆分为动车和拖车，所述列车动力分配方法包括：
    根据所述列车的行驶规划曲线、行驶线路信息及各所述动车的车辆状态，以最小纵向冲击为目标，将列车总动力分配到各所述动车；
    根据各所述动车的多个动力单元的状态，以最佳变流器系统状态为目标，将分配到本动车的给定动力进一步分配到本动车的各所述动力单元；以及
    根据各所述动力单元的轮轨粘着状态，最大化地执行分配到本动力单元的给定动力。
  2. 如权利要求1所述的列车动力分配方法，其特征在于，所述行驶规划曲线包括规划速度曲线及规划动力曲线，用于指示所述列车在行驶线路的各时刻的列车速度及列车总动力，所述行驶线路信息包括所述列车在当前时刻的坡道坡度及曲线半径，各所述动车的车辆状态包括各所述动车反馈的动车最大允许动力，将所述列车总动力分配到各所述动车的步骤包括：
    以列车总动力分配方案为求解对象，根据所述行驶规划曲线、所述行驶线路信息及各所述动车的车辆状态，对所述列车的纵向冲击进行量化建模，其中，所述列车总动力分配方案指示分配到各所述动车的动力；以及
    对构建的纵向冲击量化模型进行最优化求解，以获取最小纵向冲击对应的列车总动力分配方案。
  3. 如权利要求2所述的列车动力分配方法，其特征在于，对所述列车的纵向冲击进行量化建模的步骤包括：
    根据所述行驶规划曲线、所述行驶线路信息及各所述动车的车辆状态，计算各所述车辆之间的车钩力及车钩力冲量；以及
    根据各所述车辆之间的最大车钩力及最大车钩力冲量量化所述列车的纵向冲击，以构建所述纵向冲击量化模型。
  4. 如权利要求2所述的列车动力分配方法，其特征在于，对所述纵向冲击量化模型进行最优化求解的步骤包括：
    采用控制变量参数化方法或群体智能算法，在各所述动车的单周期动力最大允许变化量的范围内，对所述列车总动力分配方案进行最优化求解，其中，所述动车的单周期动力最大允许变化量由所述动车的车辆速度和/或变流器系统状态决定。
  5. 如权利要求2所述的列车动力分配方法，其特征在于，还包括：
    先根据列车的动力单周期最大允许变化量，对所述规划动力曲线的各时刻的列车总动力进行限幅滤波处理，其中，所述列车的动力单周期最大允许变化量由列车速度、列车网压和/或行驶线路条件决定；以及
    再根据所述限幅滤波处理后的规划动力曲线，对所述列车的纵向冲击进行量化建模。
  6. 如权利要求5所述的列车动力分配方法，其特征在于，还包括：
    根据各所述动车的动车最大允许动力计算对应时刻的列车最大允许动力；
    响应于所述列车最大允许动力小于所述限幅滤波处理后的规划动力曲线的对应时刻的列车总动力，将由各所述动车的动车最大允许动力构成的列车动力分配方案代入所述纵向冲击量化模型，以计算对应的列车冲击量化值；
    响应于所述列车冲击量化值小于量化门槛值，根据各所述动车的动车最大允许动力分配所述列车总动力，其中，所述量化门槛值是根据列车运行安全评估得到的最大允许纵向冲击量化值决定；以及
    响应于所述列车冲击量化值大于或等于所述量化门槛值，逐步减小所述对应时刻的列车总动力，直到所述列车冲击量化值小于所述量化门槛值。
  7. 如权利要求1所述的列车动力分配方法，其特征在于，所述动车的车辆状态包括本动车的各所述动力单元反馈的轮轨粘着状态系数、单元最大允许动力、电机综合转速和/或电机综合温度，将分配到本动车的给定动力进一步分配到本动车的各所述动力单元的步骤包括：
    响应于本动车的动车最大允许动力大于或等于分配到本动车的给定动力，以动车动力分配方案为求解对象，根据本动车的所述车辆状态对变流器系统状态进行量化建模，其中，所述动车动力分配方案指示分配到本动车的各所述动力单元的动力，所述动车最大允许动力是根据各所述动力单元的单元最大允许动力计算；以及
    在本动车的各所述动力单元的单周期动力最大允许变化量的范围内，对构建的系 统状态量化模型进行最优化求解，以获取最佳变流器系统状态对应的动车动力分配方案，其中，所述动力单元的单周期动力最大允许变化量由所述动车的车辆速度和/或变流器系统状态决定。
  8. 如权利要求7所述的列车动力分配方法，其特征在于，将分配到本动车的给定动力进一步分配到本动车的各所述动力单元的步骤还包括：
    响应于本动车的动车最大允许动力小于分配到本动车的给定动力，以动车动力分配方案为求解对象，根据本动车的动力发挥目标函数及变流器状态目标函数，对本动车的动力发挥及变流器系统状态进行综合的量化建模，其中，所述动力发挥目标函数指示分配到本动车的各所述动力单元的动力之和，所述变流器状态目标函数指示变流器系统状态的量化值；
    在本动车的各所述动力单元的单周期动力最大允许变化量的范围内，对构建的综合量化模型进行最优化求解，以获取最优综合情况对应的最优集前沿面；以及
    根据最低牵引力限制准则、最低变流器状态限制准则及牵引力发挥与变流器状态优先级，从所述最优集前沿面的多个动车动力分配方案中选取对应的最优解。
  9. 如权利要求7所述的列车动力分配方法，其特征在于，将分配到本动车的给定动力进一步分配到本动车的各所述动力单元的步骤还包括：
    响应于分配到本动车的给定动力小于动力门槛值，将分配到本动车的给定动力集中分配到本动车的部分动力单元，其中，所述动力门槛值是根据本动车的各所述动力单元的能量效率决定。
  10. 如权利要求1所述的列车动力分配方法，其特征在于，最大化地执行分配到本动力单元的给定动力的步骤包括：
    根据轮对的蠕滑速度及轮对加速度指标，计算本动力单元的轮轨粘着状态系数；
    响应于所述轮轨粘着状态系数指示所述轮对无空转滑行趋势，将分配到本动力单元的给定动力作为粘着给定力下发到逆变控制器；
    响应于所述轮轨粘着状态系数指示所述轮对有空转滑行趋势或已发生空转滑行，根据所述动力单元的有效轮径、轮对转动惯量、轮对速度及电机实际发挥力计算轮对粘着力，并将所述轮对粘着力作为粘着给定力下发到所述逆变控制器；以及
    以所述逆变控制器控制本动力单元的牵引电机执行所述粘着给定力。
  11. 如权利要求10所述的列车动力分配方法，其特征在于，还包括：
    将本动力单元的所述轮轨粘着状态系数反馈到对应动车的车辆级控制器，以用于组成所述对应动车的车辆状态；
    响应于所述轮轨粘着状态系数指示所述轮对无空转滑行趋势，根据本动力单元轴端当前转速下的最大允许动力、前一时刻的给定动力及所述轮轨粘着状态系数计算当前时刻的轮对粘着力，并将所述当前时刻的轮对粘着力反馈到所述车辆级控制器，以用于组成所述对应动车的车辆状态；以及
    响应于所述轮轨粘着状态系数指示所述轮对有空转滑行趋势或已发生空转滑行，选取一个完整空转滑行控制周期内所述轮对粘着力的最大值反馈到所述车辆级控制器，以用于组成所述对应动车的车辆状态。
  12. 如权利要求10所述的列车动力分配方法，其特征在于，最大化地执行分配到本动力单元的给定动力的步骤还包括：
    采集本动力单元的牵引电机的温度、电流、电压及转速，以判定本动力单元的牵引逆变状态；
    响应于所述牵引逆变状态良好，控制所述牵引电机执行所述粘着给定力；以及
    响应于所述牵引逆变状态不佳，根据所述牵引逆变状态对所述牵引电机的功率进行限制并计算对应的限功动力，对所述粘着给定力与所述限功动力进行比较，以控制牵引电机执行其中的较小值。
  13. 如权利要求12所述的列车动力分配方法，其特征在于，所述动力单元包括至少一根控制轴，所述动力分配方法还包括：
    对各所述牵引电机的转速取平均值以作为所述动力单元的电机综合转速，将所述电机综合转速反馈到对应动车的车辆级控制器，以用于组成所述对应动车的车辆状态；
    对各所述牵引电机的温度取平均值以作为所述动力单元的电机综合温度，将所述电机综合温度反馈到所述车辆级控制器，以用于组成所述对应动车的车辆状态；以及
    将所述粘着给定力与所述限功动力中的较小值作为所述动力单元的单元最大允许动力，并将所述单元最大允许动力反馈到所述车辆级控制器，以用于组成所述对应动车的车辆状态。
  14. 如权利要求1所述的列车动力分配方法，其特征在于，还包括：
    从所述列车的自动驾驶系统获取所述行驶规划曲线；
    从列车运行监控记录装置获取所述行驶线路信息；以及
    从各所述动车的车辆级控制器获取各所述动车的车辆状态，其中，各所述车辆状态分别由对应动车的多个动力单元反馈到所述车辆级控制器的单元状态组合而成。
  15. 一种列车动力分配装置，其特征在于，所述列车包括多节车辆，所述多节车辆分为动车和拖车，所述列车动力分配装置包括存储器及处理器，所述处理器连接所述存储器，并配置用于实施如权利要求1～14中任一项所述的列车动力分配方法。
  16. 一种计算机可读存储介质，其上存储有计算机指令，其特征在于，所述计算机指令被处理器执行时，实施如权利要求1～14中任一项所述的列车动力分配方法。
PCT/CN2020/130343 2020-10-29 2020-11-20 一种列车动力分配方法及装置 WO2022088316A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU2020474272A AU2020474272A1 (en) 2020-10-29 2020-11-20 Train power distribution method and device

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011179889.6A CN112249043B (zh) 2020-10-29 2020-10-29 一种列车动力分配方法及装置
CN202011179889.6 2020-10-29

Publications (1)

Publication Number Publication Date
WO2022088316A1 true WO2022088316A1 (zh) 2022-05-05

Family

ID=74261456

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/130343 WO2022088316A1 (zh) 2020-10-29 2020-11-20 一种列车动力分配方法及装置

Country Status (3)

Country Link
CN (1) CN112249043B (zh)
AU (1) AU2020474272A1 (zh)
WO (1) WO2022088316A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115546928A (zh) * 2022-11-25 2022-12-30 新誉轨道交通科技有限公司 一种车辆空转滑行辨识方法、装置、设备及存储介质
CN116476871A (zh) * 2023-04-27 2023-07-25 湖南工业大学 一种多轴电力机车牵引力平衡控制系统

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112883322B (zh) * 2021-03-10 2022-09-27 资阳中车电力机车有限公司 一种齿轨列车粘着与齿轨动力分配计算方法
CN114633780B (zh) * 2021-12-30 2022-11-25 中南大学 重载列车及其纵向动力学牵引运行优化控制系统
CN115416696B (zh) * 2022-11-04 2023-02-07 中铁工程服务有限公司 一种动力分散式电机车的行车控制方法
CN117842109A (zh) * 2023-04-27 2024-04-09 湖南工业大学 一种机车轮对空转综合判断处理装置
CN116902013B (zh) * 2023-09-11 2024-01-16 北京城建智控科技股份有限公司 列车多轴协同黏着控制方法、装置、电子设备及存储介质

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006304464A (ja) * 2005-04-19 2006-11-02 Hitachi Ltd 列車運転制御システム、統括制御装置および駆動装置
CN104960526A (zh) * 2015-07-09 2015-10-07 南车株洲电力机车研究所有限公司 一种动力分散型列车牵引力分配方法及系统
CN105015524A (zh) * 2015-07-09 2015-11-04 南车株洲电力机车研究所有限公司 一种多辆编组列车制动力分配方法及系统
CN106696760A (zh) * 2017-01-19 2017-05-24 北京理工大学 一种轮毂电机驱动车辆的动力分配方法
CN109094547A (zh) * 2018-07-27 2018-12-28 中车唐山机车车辆有限公司 城铁车辆制动力分配方法及装置
CN109532963A (zh) * 2019-01-15 2019-03-29 西南交通大学 一种电力列车的经济性动力匹配与分配方法
CN110091847A (zh) * 2018-01-29 2019-08-06 湖南工业大学 一种高速列车制动力优化分配方法
CN110333655A (zh) * 2019-07-02 2019-10-15 北京交通大学 一种高速列车多动力单元分布式控制方法

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101079903B1 (ko) * 2009-08-24 2011-11-04 엘에스산전 주식회사 열차자동운전에서의 속도 제어장치 및 방법
US9337923B2 (en) * 2011-10-18 2016-05-10 Nec Corporation Optical spectral shaping for nonlinearity enhancement in high speed digital coherent optical transmissions
JP6591316B2 (ja) * 2016-03-09 2019-10-16 株式会社東芝 監視制御装置
KR102134308B1 (ko) * 2017-09-28 2020-07-15 한국철도기술연구원 중련 운행 방식의 전기 기관차 무선 분산 제어 시스템 및 방법
CN110654420A (zh) * 2018-06-29 2020-01-07 株洲中车时代电气股份有限公司 一种列车恒速控制方法及控制系统
CN108995664B (zh) * 2018-07-25 2019-12-27 中车大连机车车辆有限公司 列车的动力调整方法和装置

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006304464A (ja) * 2005-04-19 2006-11-02 Hitachi Ltd 列車運転制御システム、統括制御装置および駆動装置
CN104960526A (zh) * 2015-07-09 2015-10-07 南车株洲电力机车研究所有限公司 一种动力分散型列车牵引力分配方法及系统
CN105015524A (zh) * 2015-07-09 2015-11-04 南车株洲电力机车研究所有限公司 一种多辆编组列车制动力分配方法及系统
CN106696760A (zh) * 2017-01-19 2017-05-24 北京理工大学 一种轮毂电机驱动车辆的动力分配方法
CN110091847A (zh) * 2018-01-29 2019-08-06 湖南工业大学 一种高速列车制动力优化分配方法
CN109094547A (zh) * 2018-07-27 2018-12-28 中车唐山机车车辆有限公司 城铁车辆制动力分配方法及装置
CN109532963A (zh) * 2019-01-15 2019-03-29 西南交通大学 一种电力列车的经济性动力匹配与分配方法
CN110333655A (zh) * 2019-07-02 2019-10-15 北京交通大学 一种高速列车多动力单元分布式控制方法

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115546928A (zh) * 2022-11-25 2022-12-30 新誉轨道交通科技有限公司 一种车辆空转滑行辨识方法、装置、设备及存储介质
CN115546928B (zh) * 2022-11-25 2023-01-31 新誉轨道交通科技有限公司 一种车辆空转滑行辨识方法、装置、设备及存储介质
CN116476871A (zh) * 2023-04-27 2023-07-25 湖南工业大学 一种多轴电力机车牵引力平衡控制系统
CN116476871B (zh) * 2023-04-27 2024-05-28 湖南工业大学 一种多轴电力机车牵引力平衡控制系统

Also Published As

Publication number Publication date
AU2020474272A1 (en) 2023-06-22
CN112249043B (zh) 2022-06-10
CN112249043A (zh) 2021-01-22

Similar Documents

Publication Publication Date Title
WO2022088316A1 (zh) 一种列车动力分配方法及装置
CN109783890B (zh) 基于钩缓装置模型的重载列车运行曲线多目标优化方法
CN110525487A (zh) 一种基于车钩力约束的自动驾驶方法及系统
TWI490139B (zh) 軌道列車節省電能運行控制方法
CN104881527A (zh) 城市轨道交通列车ato速度命令优化方法
CN110549868A (zh) 基于动力系统实时功率的混合动力有轨电车速度调整方法
WO2023126023A1 (zh) 重载列车及其纵向动力学牵引运行优化控制系统、方法
Hillmansen et al. Electric railway traction systems and techniques for energy saving
CN108376190B (zh) 一种确定城轨列车车载超级电容器组参数的方法
CN110955146A (zh) 基于分布式模型预测控制的重载机车粘着控制方法及装置
CN113821966A (zh) 高速磁浮列车运行节能优化方法、系统及存储介质
CN109703593B (zh) 一种无接触网有轨电车全区间运行综合优化方法
CN109376437A (zh) 一种基于多目标遗传算法的新能源车载储能优化控制方法
CN112109773B (zh) 一种基于h∞和多目标优化的列车速度控制方法
Schenker et al. Optimization model for operation of battery multiple units on partly electrified railway lines
Yu et al. Research on operating strategy based on particle swarm optimization for heavy haul train on long down-slope
Gu et al. Energy efficient driving strategy for trains in MRT systems
Albrecht et al. Using maximum power to save energy
CN109572451B (zh) 一种混合动力有轨电车的充电方法及其仿真计算方法
CN104890530B (zh) 对轮对的不均匀磨损进行修正的方法
Sarsembayev et al. Analyze of impact of track alignment on the energy consumption level
Ke et al. Model of traction system and speed control for single train of Taipei mass rapid transit system
Su et al. Optimization of cyclic air braking strategy for heavy haul trains: an ADP approach
Wu et al. Simulation of heavy haul train energy consumption with locomotive adhesion model
CN106240582B (zh) 能发电的火车多功能车厢

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20959475

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2020474272

Country of ref document: AU

Date of ref document: 20201120

Kind code of ref document: A

122 Ep: pct application non-entry in european phase

Ref document number: 20959475

Country of ref document: EP

Kind code of ref document: A1