A high-efficiency electric control method and system for a traction motor of a new energy electric locomotive

Abstract

The present application belongs to the field of electric traction control, and particularly relates to a high-efficiency electric control method and system for traction motors of new energy electric locomotives, which comprises: operation state parameter acquisition; bus voltage feedforward dynamic compensation; online identification of key parameters of electric machines; multi-target coordinated optimization traction control; active damping injection resonance suppression; fault diagnosis and fault-tolerant control. The method aims to solve the problem of control accuracy decline caused by voltage fluctuation and parameter drift. At the same time, a system is constructed based on the above method, which comprises: an energy management interface unit; a signal conditioning and acquisition subsystem; a main control core processing unit; an isolation driving subsystem; a power conversion module; an energy feedback management unit; a residual life prediction module. By adopting the above technical scheme, the control stability and energy efficiency ratio under complex working conditions are significantly improved, the loss and torque ripple are reduced, and the smoothness and safety of locomotive operation are enhanced.

Description

Technical Field

[0001] This invention belongs to the field of electric traction control, specifically relating to a high-efficiency electronic control method and system for traction motors in new energy electric locomotives. Background Technology

[0002] With the advancement of global energy structure transformation and the goal of decarbonizing rail transportation, new energy electric locomotives are being used more and more widely in mining, ports, and short-distance freight transport. In these locomotives, the traction motor and its electronic control system are the core components that determine the overall range, power response, and safety of the vehicle.

[0003] Currently, most mainstream new energy electric locomotive traction control systems employ vector control or direct torque control technologies based on space vector pulse width modulation. These technologies can achieve effective speed and torque regulation under steady-state conditions through field orientation and current decoupling.

[0004] However, existing technologies have the following significant limitations when faced with the complex operating conditions unique to new energy locomotives:

[0005] 1. Poor adaptability to dynamic fluctuations in bus voltage: Unlike the constant voltage of traditional overhead contact lines, the DC bus voltage of new energy locomotives is highly dynamic and time-varying, influenced by battery state of charge, internal resistance, and high-power charging and discharging. Especially during heavy-load starting, rapid acceleration and deceleration, or regenerative braking, large fluctuations in bus voltage can interfere with the inverter's modulation linearity. Existing linear control methods based on fixed-parameter proportional-integral regulators are insufficient to effectively suppress such disturbances, easily leading to current loop decoupling failure, torque pulsation, reduced energy efficiency, and impact on the mechanical transmission chain.

[0006] 2. There is a contradiction between efficiency improvement and system stability: In order to reduce harmonic distortion rate and reduce losses, existing technologies often try to increase the switching frequency or use complex algorithms. However, this will significantly increase the inverter switching losses and place extremely high demands on the computing power of the controller. In high-frequency vibration environments, the system is prone to oscillation due to control delay.

[0007] 3. Lack of adaptive compensation for parameter drift: Temperature rise during traction motor operation can cause nonlinear shifts in key parameters such as stator resistance and rotor reactance. Existing solutions typically lack adaptive compensation mechanisms for real-time changes in these parameters, causing the control algorithm to deviate from the theoretically optimal efficiency point across the entire temperature and speed domains.

[0008] In summary, existing traction motor control technologies suffer from drawbacks when dealing with the highly nonlinear, time-varying parameters, and highly coupled dynamic systems of new energy electric locomotives. These shortcomings include a single control dimension, insufficient disturbance rejection capability, and significant bottlenecks in energy efficiency improvement. Therefore, there is an urgent need to develop a high-efficiency electronic control method and system for traction motors that can deeply integrate the energy characteristics of new energy sources, possess strong disturbance rejection capability, and achieve optimal global energy efficiency. Summary of the Invention

[0009] The primary objective of this invention is to provide an efficient electronic control method for traction motors in new energy electric locomotives. This method aims to address the technical problems of reduced system control accuracy, decreased operating efficiency, and sluggish power response caused by significant fluctuations in the DC bus voltage output from energy sources such as power batteries or supercapacitors, nonlinear drift of motor physical parameters with temperature rise, and technical contradictions between inverter switching losses and harmonic suppression during the operation of new energy electric locomotives.

[0010] The second objective of this invention is to provide a high-efficiency electronic control system for traction motors in new energy electric locomotives.

[0011] To achieve the above objectives, the technical solution adopted by the present invention is as follows: a high-efficiency electronic control method for traction motors of new energy electric locomotives, comprising the following specific steps:

[0012] S1. Acquisition of operating status parameters: The operating status parameters of the new energy electric locomotive are acquired in real time through an airborne sensor array; the operating status parameters include DC bus voltage, three-phase stator current of traction motor, rotor mechanical position angle, rotor speed and stator winding temperature;

[0013] S2. Bus voltage feedforward dynamic compensation: Construct a bus voltage feedforward dynamic compensation model, receive the instantaneous value of DC bus voltage and compare it with the preset nominal bus voltage reference value, and calculate the voltage fluctuation deviation rate.

[0014] S3. Online identification of key motor parameters: For the stator resistance temperature rise drift and rotor permanent magnet flux thermal decay characteristics of traction motor under heavy load operation, online parameter identification based on multi-physics field coupling characteristics is performed.

[0015] S4. Multi-objective collaborative optimization traction control: Construct a hierarchical multi-objective collaborative optimization control architecture, based on the maximum torque-current ratio control, and obtain the initial optimal current vector command by combining the pre-stored motor efficiency map online lookup table;

[0016] S5. Active damping injection resonance suppression: To address the electromechanical coupling resonance problem caused by the complex mechanical transmission chain and long cable connection of new energy electric locomotives, an active damping feedback term is superimposed at the output end of the speed regulator.

[0017] S6. Fault Diagnosis and Fault-Tolerant Control: Real-time monitoring of key status parameters of the power conversion module, including power device junction temperature, DC bus voltage fluctuation rate, three-phase current imbalance, and on-state voltage drop characteristics of power devices; construction of a multi-level fault diagnosis and fault-tolerant decision-making mechanism.

[0018] Preferably, in step S1, the DC bus voltage is acquired by a Hall closed-loop voltage sensor based on the magnetic compensation principle, and the current loop control gain is dynamically adjusted based on the fluctuation rate of the acquired signal; the three-phase stator current is acquired by a through-hole low-drift current transformer, and the fundamental signal is obtained by synchronous sampling technology synchronized with the extreme point of the pulse width modulation carrier, and then decomposed into excitation and torque components by coordinate transformation; the rotor mechanical position angle and rotor speed are acquired by an absolute value photoelectric encoder with differential signal output function, and a sensorless control backup scheme based on a sliding mode observer is configured to automatically switch to estimation mode when the main encoder fails; the stator winding temperature is monitored in real time by a thermal resistance sensor embedded in the stator winding.

[0019] Preferably, in step S2, the voltage fluctuation deviation rate is introduced into the modulation index correction link of the space vector pulse width modulator to adjust the conduction time of the inverter switching transistor in real time, so as to offset the influence of bus voltage fluctuation on the equivalent output voltage vector. The compensation model is equipped with a voltage safety window and amplitude limiting protection logic. When the voltage fluctuation exceeds the safety window threshold, the correction amplitude of the modulation index is automatically limited to prevent the current waveform distortion caused by entering the deep over-modulation zone, thereby maintaining the excitation current and torque current at the stator end of the motor in a decoupled state under voltage fluctuation environment. When processing the voltage signal, the bus voltage feedforward dynamic compensation model uses a sliding window weighted average filtering algorithm to remove non-characteristic voltage spikes caused by unstable pantograph current collection or internal impedance switching of the power battery.

[0020] Preferably, the specific operation process of step S3 is as follows: Collect stator winding temperature signals and current loop residual signals, construct an adaptive observer, and dynamically update the stator resistance and rotor flux linkage parameters in the prediction model by comparing the current estimation value output by the motor prediction model with the measured current value, using an error feedback mechanism; the online parameter identification adopts a time-sharing identification logic adapted to operating conditions: during the locomotive's steady-state cruise phase, the rotor flux linkage parameters are locked, and online identification of the stator resistance is prioritized; during high-current dynamic switching phases such as locomotive acceleration or braking, the stator resistance parameters are locked, and online calibration of the rotor flux linkage is prioritized; furthermore... The adaptive observer also integrates a high-frequency signal injection auxiliary mechanism: when the motor is running at low speed, a high-frequency disturbance signal is injected into the stator winding, and the rotor flux linkage and inductance parameters are extracted based on the motor's impedance response to the signal; the real-time updated stator resistance and rotor flux linkage parameters are smoothed and filtered before being fed back to the current loop controller to eliminate torque control deviations caused by parameter mismatch; wherein, the online parameter identification adopts a time-division identification strategy: during the locomotive's steady-state cruise phase, online identification of the stator resistance is prioritized; during the high-current dynamic switching phase of the locomotive's acceleration or braking, online calibration of the rotor flux linkage is prioritized.

[0021] The identification process also introduces an identification method based on injected signals: during motor operation, a high-frequency non-characteristic current vector is injected into the stator current, the frequency of which is higher than the motor's mechanical response frequency.

[0022] Preferably, in step S4, the hierarchical multi-objective collaborative optimization control architecture is generated based on the initial instructions of the efficiency map; it integrates a nonlinear magnetic saturation compensation mechanism: a three-dimensional inductance mapping table considering the coupling effect of direct-axis and quadrature-axis currents is established, and the accurate inductance value is obtained by online interpolation based on the current current vector coordinates in real-time control, and the proportional-integral gain of the current loop is dynamically adjusted accordingly to eliminate the dynamic performance degradation caused by magnetic saturation; adaptive field weakening and copper loss optimization: under low-speed, high-torque conditions, the stator current vector phase angle is finely adjusted based on the real-time identified inductance parameters to minimize stator copper losses; under high-speed, constant-power conditions, it automatically switches to the field weakening control mode, and expands the speed regulation range by increasing the negative direct-axis current component to offset the back electromotive force; the inverter carrier frequency is dynamically adjusted according to the constraints of speed level and load rate; the carrier frequency is increased under heavy-load, low-speed conditions to suppress torque ripple, and the carrier frequency is decreased under high-speed, stable conditions to reduce switching losses.

[0023] Preferably, in step S5, the active damping feedback term is generated by extracting the mechanical resonant frequency component from the acquired rotor speed signal through bandpass or high-pass filtering, and then multiplying it by an adaptive damping coefficient; the cutoff frequency of the filter is set to dynamically change with the mechanical resonant frequency of the system, and the phase of the generated damping feedback term is opposite to the direction of the mechanical vibration velocity; the active damping feedback term is injected into the torque current setpoint, and the reverse electromagnetic torque generated by the motor actively absorbs the vibration energy of the mechanical transmission chain and cable system, suppresses the resonance peak of the drive system, and improves the smoothness of locomotive operation.

[0024] Preferably, in step S6, the multi-level fault diagnosis and fault-tolerant decision-making mechanism is as follows: when sensor drift, power device performance degradation, or slight DC bus overvoltage is detected, it is determined to be a level one fault, and the system automatically switches to emergency derating operation mode. By linearly limiting the maximum output torque and dynamically reducing the inverter switching frequency, thermal stress is reduced and the locomotive is ensured to travel slowly to the maintenance station; when power device open circuit, severe overcurrent, or severe DC bus overvoltage is detected, it is determined to be a level two fault, and a control reconfiguration or safety shutdown strategy is triggered. The faulty bridge arm is blocked and the system switches to the non-faulty phase to maintain operation or performs inertial coasting stop.

[0025] Furthermore, the present invention also provides a high-efficiency electronic control system for the traction motor of a new energy electric locomotive, for realizing the aforementioned high-efficiency electronic control method for the traction motor of the new energy electric locomotive, comprising:

[0026] The energy management interface unit is used to connect the power battery and the braking resistor to perform bidirectional energy flow scheduling and bus voltage stabilization control.

[0027] The signal conditioning and acquisition subsystem is used to acquire the three-phase current, bus voltage and temperature signals of the traction motor in real time, and to perform filtering and analog-to-digital conversion on the signals.

[0028] The main control core processing unit is communicatively connected to the energy management interface unit and the signal conditioning and acquisition subsystem, respectively, and is used to execute traction control algorithms, online parameter identification, fault diagnosis, and multi-objective optimization decision-making. The main control core processing unit adopts a heterogeneous parallel computing architecture composed of a digital signal processor (DSP) and a field-programmable gate array (FPGA). The DSP and FPGA exchange data through a high-speed parallel bus. The DSP is configured to execute advanced control logic, including bus voltage feedforward calculation, online motor parameter identification, multi-objective collaborative optimization algorithms, and fault diagnosis logic. The FPGA is configured to handle high-frequency and high real-time tasks, including synchronous trigger sampling of multiple analog signals, hard decoding of encoder signals, generation of space vector pulse width modulation waveforms, and dead time compensation.

[0029] An isolated drive subsystem, connected to the main control core processing unit, is used to convert control signals into drive signals and transmit them with electrical isolation.

[0030] A power conversion module, connected to the isolated drive subsystem, is used to invert DC power into AC power to drive the traction motor.

[0031] The energy feedback management unit is configured to dynamically allocate electric braking power according to the allowable charging power and state of charge of the power battery when the locomotive is braking, and automatically activate the braking resistor branch when the battery pack cannot absorb all the braking energy to consume excess energy through pulse width modulation.

[0032] The remaining life prediction module is configured to assess the health status of core components online by recording the historical junction temperature cycle number of the power module, the ripple current load of the capacitor, and the cumulative operating temperature duration of the motor, combined with a preset loss degradation model, and transmit the prediction results to the ground operation and maintenance center.

[0033] Preferably, the signal conditioning and acquisition subsystem includes multiple synchronous sampling channels, each of which is sequentially configured with an anti-aliasing filter, a precision instrumentation amplifier, and an analog-to-digital converter; the anti-aliasing filter adopts a fourth-order Butterworth structure, and its cutoff frequency is set according to twice the switching frequency of the power conversion module; the current sampling channel adopts a differential input method and is connected to the current transformer through shielded twisted-pair cable;

[0034] The isolated drive subsystem employs a drive chip based on fiber optic communication or magnetic isolation technology, featuring dynamic Miller clamping and desaturation protection circuitry. The desaturation protection circuitry is configured to monitor the on-state voltage drop of the power transistor in real time, and upon detecting an overcurrent condition, to shut down the trigger signal within 2μs and issue a hardware interrupt to the main control core processing unit. The isolated drive subsystem also integrates active gate voltage control technology, configured to dynamically adjust the gate current during the transistor's on-state phase via multi-stage resistor switching, and to suppress voltage spikes and reduce turn-off losses during the turn-off phase by controlling the discharge current magnitude.

[0035] The power conversion module uses high-power-density silicon carbide MOSFETs or IGBTs, and integrates a temperature sensor closely attached to the power chip. The power conversion module is mounted on a heat sink with a forced air cooling or liquid cooling structure, and its AC output terminal is equipped with a sine wave filter or dVdt filter composed of a high-voltage inductor and a non-polar thin-film capacitor. The DC side of the power conversion module adopts a stacked bus structure and is equipped with a high-frequency, long-life thin-film supporting capacitor. The AC output port is equipped with a high-frequency current transformer to monitor the common-mode current and cut off the inverter output when the leakage current exceeds a preset threshold.

[0036] Preferably, the main control core processing unit further includes a historical data storage module, which uses a highly reliable ferroelectric random access memory. The historical data storage module is configured to record all key control parameter waveforms within one minute before the locomotive failure occurs, and to lock and report the data as black box information when the system experiences an abnormal shutdown. The remaining life prediction module works in conjunction with the historical data storage module to use the fault waveform data recorded by the black box to correct the loss degradation model, so as to achieve full life cycle operating cost optimization and preventive maintenance.

[0037] The technical solution provided by this invention has the following beneficial effects:

[0038] First, by introducing a bus voltage feedforward dynamic compensation model, this invention significantly improves the control stability of new energy electric locomotives under conditions of large bus voltage fluctuations. Because it can correct the inverter's duty cycle in real time, the system eliminates torque drop caused by bus voltage dips, ensuring the locomotive's power continuity under extreme conditions such as heavy-load starting, and reducing torque ripple rate by more than 30% compared to traditional control schemes.

[0039] Secondly, the online identification mechanism for motor parameters based on the multi-physics coupling characteristics effectively solves the control point deviation problem caused by the temperature rise of the traction motor. By dynamically updating the stator resistance and flux linkage parameters, the decoupling accuracy of the current loop is significantly improved, resulting in a 1.5–2.5% increase in the energy conversion efficiency of the motor across the entire temperature range, and a significant extension of the driving range of new energy locomotives.

[0040] Furthermore, the multi-objective collaborative optimization control logic achieves a global balance between switching losses and harmonic characteristics. Through dynamic adjustment of the carrier frequency, the system can significantly reduce the temperature rise of power devices in the high-efficiency demand range, while providing smooth current output in the low-speed stability demand range, thereby reducing the electromagnetic noise and temperature rise rate of the traction motor and improving the comfort of locomotive operation.

[0041] Furthermore, this invention significantly reduces the low-order harmonic content in the output voltage through real-time compensation of the inverter's dead time. Because power transistors have tail currents during turn-off, traditional fixed dead times cause deviations in the phase and amplitude of the voltage vector, especially noticeable at low frequencies. This invention detects the polarity of the output current and calculates the turn-on and turn-off delays of the power transistors in real-time within the FPGA, dynamically compensating for the pulse width to ensure the output voltage vector closely follows the given vector. This technique effectively suppresses the fifth and seventh harmonics in the stator current, reduces motor torque ripple and additional iron losses, and enables the locomotive to maintain stable power output even at extremely low speeds.

[0042] Furthermore, this invention integrates idling and coasting control functions into the traction control logic. By comparing the speed change rate of each axle motor and the deviation from the locomotive radar speed measurement value in real time, the system can quickly identify the initial idling state of the wheelset. Once the risk of idling is detected, the system actively adjusts the traction force by reducing the torque current setpoint, and then smoothly restores the torque after the adhesion performance is restored. This highly sensitive adhesion control greatly improves the traction efficiency utilization rate of new energy locomotives under wet and slippery track conditions.

[0043] Meanwhile, the system incorporates a multi-level overcurrent protection mechanism at the logic layer. The first layer is an FPGA-based hardware instantaneous current cutoff protection with a response time of less than 1μs, directly shutting off the power transistor drive signal. The second layer is a DSP-based peak current limiting protection, which introduces dynamic limiting within the current loop to prevent current overshoot caused by sudden torque command changes. The third layer is an inverse-time overcurrent protection based on a thermal model, simulating the heat accumulation process of the motor windings, allowing the motor to operate under overload for a short period to provide additional starting torque, while ensuring that the winding temperature does not exceed the insulation class limit. This layered protection strategy balances the system's safety and reliability with its dynamic overload capability, meeting the complex operating conditions required by railway traction.

[0044] Moreover, the active damping injection technology and highly reliable hardware architecture design greatly enhance the system's anti-disturbance capability and service life. By actively suppressing mechanical resonance, fatigue damage to the transmission gearbox and coupling is reduced; while the highly integrated isolated drive and fault diagnosis logic provide a solid hardware guarantee for the safe operation of the locomotive. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a flowchart illustrating a high-efficiency electronic control method for traction motors in new energy electric locomotives according to the present invention.

[0047] Figure 2 This is a schematic diagram of the overall structure of a high-efficiency electronic control system for traction motors in new energy electric locomotives according to the present invention. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] like Figure 1 As shown, the technical solution adopted by the present invention is: a high-efficiency electronic control method for traction motors of new energy electric locomotives, comprising the following specific steps:

[0050] S1. Acquisition of operating status parameters: The operating status parameters of the new energy electric locomotive are acquired in real time through the airborne sensor array. The operating status parameters include DC bus voltage, three-phase stator current of traction motor, rotor mechanical position angle, rotor speed and stator winding temperature.

[0051] S2. Bus voltage feedforward dynamic compensation: Construct a bus voltage feedforward dynamic compensation model, receive the instantaneous value of DC bus voltage and compare it with the preset nominal bus voltage reference value, and calculate the voltage fluctuation deviation rate.

[0052] S3. Online identification of key motor parameters: For the stator resistance temperature rise drift and rotor permanent magnet flux thermal decay characteristics of traction motor under heavy load operation, online parameter identification based on multi-physics field coupling characteristics is performed.

[0053] S4. Multi-objective collaborative optimization traction control: Construct a hierarchical multi-objective collaborative optimization control architecture, based on maximum torque-current ratio control, and obtain the initial optimal current vector command by combining the pre-stored motor efficiency map online lookup table.

[0054] S5. Active damping injection resonance suppression: To address the electromechanical coupling resonance problem caused by the complex mechanical transmission chain and long cable connection of new energy electric locomotives, an active damping feedback term is superimposed at the output end of the speed regulator.

[0055] S6. Fault Diagnosis and Fault-Tolerant Control: Real-time monitoring of key status parameters of the power conversion module, including power device junction temperature, DC bus voltage fluctuation rate, three-phase current imbalance, and on-state voltage drop characteristics of power devices; construction of a multi-level fault diagnosis and fault-tolerant decision-making mechanism.

[0056] In the operation of acquiring operating status parameters in step S1, the operating status parameters of the new energy electric locomotive are acquired in real time through the airborne sensor array. The operating status parameters include DC bus voltage, three-phase stator current of traction motor, rotor mechanical position angle, rotor speed and stator winding temperature.

[0057] The DC bus voltage is acquired using a Hall-effect closed-loop voltage sensor based on the magnetic compensation principle. The current loop control gain is dynamically adjusted based on the fluctuation rate of the acquired signal. Specifically, the system uses a Hall-effect closed-loop voltage sensor with a frequency response bandwidth of at least 10kHz, located at the power supply output, to acquire the voltage signal and capture high-frequency ripple. The system samples the voltage signal at a frequency of 20kHz and performs real-time sliding mean filtering within the FPGA. The filtering window is set to four sampling points. Experimental verification shows that this window size can filter out switching noise while controlling the phase lag within 50μs, thus meeting the current loop bandwidth requirements. The filtered voltage value is used as the denominator in the correction coefficient calculation, directly correcting the vector action time of the space vector pulse width modulation. Specifically, the system calculates the bus voltage change rate in real time. When the fluctuation rate exceeds 5% / ms, a transient enhanced response mode is immediately activated: the system dynamically increases the proportional gain of the current loop according to a linear mapping relationship based on the absolute value of the fluctuation rate, and sets the upper limit of the gain as a critical value to prevent oscillation; when the fluctuation rate is below the threshold for three consecutive control cycles, the gain linearly drops back to the normal value within 10ms. This strategy effectively suppresses torque ripple caused by voltage surges, while avoiding system instability due to excessive gain.

[0058] The three-phase stator current is acquired using a through-type low-drift current transformer, and the fundamental signal is obtained through synchronous sampling technology synchronized with the extreme point of the pulse width modulation carrier. Subsequently, a coordinate transformation is performed to decompose it into excitation and torque components. Specifically, the three-phase stator current acquisition and decoupling process involves: acquiring the current using a through-type low-drift current transformer, which is then converted into a voltage signal by a high-precision sampling resistor. Synchronous sampling technology is used to strictly lock the trigger time of the analog-to-digital converter to the peak or valley of the PWM carrier, thereby avoiding switching spike noise and obtaining a pure fundamental current signal. The acquired three-phase current is then transformed from a stationary coordinate system to a synchronously rotating coordinate system, precisely decomposed into excitation current and torque current components, providing accurate feedback for vector control.

[0059] The rotor mechanical position angle and rotor speed are acquired by an absolute photoelectric encoder with differential signal output function, and a sensorless control backup scheme based on sliding mode observer is configured. In the event of a main encoder failure, the system automatically switches to estimation mode. The rotor position and speed detection and bumpless switching fault tolerance are specifically manifested as follows: The position and speed are acquired using an absolute photoelectric encoder integrated at the motor shaft end with differential signal output function. The differential signal effectively suppresses common-mode interference in the locomotive's strong electromagnetic environment. The FPGA internally performs quadruple frequency counting on the differential pulses to obtain high-resolution position, and calculates the instantaneous speed by calculating the time interval between adjacent pulses. Simultaneously, the system runs a sensorless estimation algorithm based on sliding mode observer in parallel in the background. During normal operation, the observer continuously tracks the encoder feedback value and calculates the phase deviation between the two. Once a communication interruption or signal loss of the main encoder is detected, the system immediately triggers the switching logic: first, the currently calculated phase deviation is used to instantaneously correct the observer's estimated angle using phase alignment; then, within 10μs, the control loop feedback source is seamlessly switched from the encoder to the observer's estimated value. This "pre-tracking + instantaneous correction" mechanism eliminates the angle jump at the moment of switching, ensuring that the traction power is not interrupted and there is no current surge.

[0060] The stator winding temperature is monitored in real time by a thermal resistance sensor embedded in the stator winding. Specifically, the temperature monitoring is performed by using the thermal resistance sensor embedded in the stator winding to monitor the winding temperature in real time. The sampled data is processed by a first-order low-pass filter and then used as the input variable for subsequent online identification of motor resistance parameters and thermal compensation calculation.

[0061] In the bus voltage feedforward dynamic compensation operation in step S2, a bus voltage feedforward dynamic compensation model is constructed to correct the inverter's output duty cycle in real time, thus solving the technical problem that traditional feedforward control is prone to system instability under extreme conditions. The voltage fluctuation deviation rate is introduced into the modulation index correction link of the space vector pulse width modulator, adjusting the on-time of the inverter's switching transistors in real time to offset the impact of bus voltage fluctuations on the equivalent output voltage vector. The compensation model has a voltage safety window and amplitude limiting protection logic. When the voltage fluctuation exceeds the safety window threshold, the correction amplitude of the modulation index is automatically limited to prevent current waveform distortion caused by entering the deep overmodulation region, thereby maintaining the decoupled state of the excitation current and torque current at the motor stator end under voltage fluctuation conditions. When processing the voltage signal, the bus voltage feedforward dynamic compensation model uses a sliding window weighted average filtering algorithm to remove non-characteristic voltage spikes caused by pantograph current instability or internal impedance switching of the power battery. The specific implementation process of the bus voltage feedforward dynamic compensation operation includes: signal preprocessing and filtering, and deviation calculation and piecewise linear adjustment.

[0062] In this process, signal preprocessing and filtering involve acquiring the instantaneous DC bus voltage value in model receiving step S1. To address high-frequency non-characteristic voltage spikes generated by pantograph offline arcing or power battery relay switching in new energy electric locomotives, the system employs a sliding window weighted average filtering algorithm for preprocessing. This algorithm assigns high weights (e.g., 0.4) to data at the center of the window and low weights (e.g., 0.1) to data at the edges. This filters out high-frequency noise while keeping the signal group delay within half a sampling period, ensuring the phase accuracy of the feedforward compensation.

[0063] In addition, deviation calculation and piecewise linear adjustment involve comparing the filtered voltage value with the preset nominal bus voltage reference value U. nom Compare and calculate the voltage fluctuation deviation rate. .

[0064] The system sets a voltage safety window [U min U max When the bus voltage fluctuation is within the safety window, the correction factor K is linearly related to the deviation rate δ. This directly affects the modulation index correction link of the space vector pulse width modulator (SVPWM). By adjusting the on-time of the inverter's switching transistors in real time, it precisely offsets voltage fluctuations, ensuring a constant equivalent output voltage vector. When the voltage exceeds the safety window threshold (e.g., a sharp drop or surge), the system immediately activates the limiting protection logic. At this time, the correction coefficient K is forcibly clamped within the preset maximum / minimum threshold range to prevent the modulation index from entering the deep overmodulation region, thus avoiding excitation and torque current decoupling failure and motor torque pulsation due to severe distortion of the output voltage waveform.

[0065] To address the voltage characteristic changes caused by increased battery internal resistance in cold environments, the model integrates a low-temperature performance correction coefficient. This coefficient is dynamically adjusted based on data from the ambient temperature sensor. Specifically, in low-temperature mode, the response sensitivity of the voltage feedforward is appropriately reduced to prevent excessive duty cycle compensation triggered by a sudden drop in battery terminal voltage, thus avoiding a "positive feedback" effect of voltage drop. Furthermore, the lower limit of the aforementioned "voltage safety window" is reduced in real-time based on temperature, activating the limiting protection in advance to prevent excessive power extraction from causing the battery voltage to momentarily break through the undervoltage protection threshold.

[0066] Step S3 performs online identification of key motor parameters based on multi-physics coupling characteristics, aiming to solve the problem of stator resistance R of traction motor under long-term high-load operation. s As temperature rises, the magnetic flux linkage ψ of the rotor permanent magnet increases. f The problem of decreased control performance caused by physical properties that weaken with increasing temperature.

[0067] First, multi-source signal acquisition and preprocessing. The system acquires the stator winding temperature signal T in real time. stator and the residual signal e of the current loop i An adaptive observer is constructed, with the residual signal being the difference between the measured current and the current estimated by the observer. The temperature signal, after being filtered by a first-order low-pass filter to remove high-frequency noise, serves as the lookup basis for the initial parameter values; the residual signal serves as the core driving input of the adaptive observer.

[0068] Second, the system employs a time-sharing identification strategy that adapts to different operating conditions. To address the strong coupling between stator resistance and rotor flux linkage in the mathematical model, the system uses a time-sharing identification strategy based on the locomotive's operating state machine execution mode. This strategy compares the estimated current value output by the motor prediction model with the measured current value, and utilizes an error feedback mechanism to dynamically update the stator resistance R in the prediction model. s and rotor flux ψ f parameter.

[0069] Specifically, the steady-state cruise mode is characterized by stable back EMF and a significant proportion of resistance voltage drop. The system locks the rotor flux linkage ψ. f Based on the identification value from the previous cycle, using the d-axis voltage equation or the DC injection principle, the stator resistance R is preferentially considered. s Online identification. In this mode, the observer gain matrix is ​​optimized for resistance sensitivity, ensuring rapid tracking of resistance changes under constant load.

[0070] Specifically, the dynamic switching mode is characterized by: rapid changes in induced electromotive force and abundant flux linkage information under high-current conditions such as acceleration or electric braking. The system locks the stator resistance R. s The rotor flux linkage ψ is prioritized based on the current temperature lookup table value or the most recent steady-state identification value. fOnline calibration is performed. In this mode, the high sensitivity of the q-axis back EMF component to the flux linkage is utilized to quickly correct the flux linkage estimate through error feedback. During the locomotive's steady-state cruise phase, the rotor flux linkage parameters are locked, and the stator resistance R is calibrated first. s Online identification; locking the stator resistance R during high-current dynamic switching phases such as locomotive acceleration or braking. s Parameters, prioritizing rotor flux linkage ψ f Online calibration.

[0071] Third, in the extremely low speed range, such as zero speed and <5% of rated speed, where there is a significant bottleneck due to the disappearance of back electromotive force, this invention introduces a high-frequency voltage signal as a complement.

[0072] First, a frequency f is superimposed on the fundamental control voltage. inj Rotating high-frequency voltage vector u inj For example, the preferred frequency is 500–1000 Hz, which is much higher than the mechanical response frequency of the motor; secondly, since the motor behaves as an inductive load at this frequency, the current response i inj The amplitude and phase of the flux linkage mainly depend on the inductance matrix of the motor and the saturation characteristics of the permanent magnet flux linkage. When the motor is running at low speed, a high-frequency disturbance signal is injected into the stator winding. Based on the motor's impedance response to this signal, the rotor flux linkage and inductance parameters are extracted. The real-time updated stator resistance and rotor flux linkage parameters are smoothed and filtered before being fed back to the current loop controller to eliminate torque control deviations caused by parameter mismatch. This is mainly achieved by extracting the high-frequency current component through a bandpass filter to calculate the rotor position and flux linkage saturation level. In the low-speed range, the observer directly uses the flux linkage value obtained by the high-frequency injection method; in the medium- and high-speed range, the value obtained by the model method is used, and a weighted smoothing algorithm is applied for transition. This method ensures that the locomotive can still obtain high-precision flux linkage parameters under low-speed, high-torque conditions such as starting and climbing, completely eliminating low-speed torque pulsation.

[0073] It should be noted that the online parameter identification adopts a time-division identification strategy: during the locomotive's steady-state cruise phase, online identification of the stator resistance is prioritized; during the high-current dynamic switching phase of the locomotive's acceleration or braking, online calibration of the rotor flux is prioritized; the identification process also introduces an identification method based on injection signals: during motor operation, a high-frequency non-characteristic current vector is injected into the stator current, and the frequency of the high-frequency non-characteristic current vector is higher than the motor's mechanical response frequency; by detecting the motor's impedance response to the injected signal, the rotor flux and inductance parameters are extracted, achieving high-precision identification under zero-speed and extremely low-speed conditions.

[0074] Fourth, the identified raw parameters are not directly used for control, but are the final parameters R after amplitude limiting and moving average filtering. final ,ψ finaThe current loop PI controller is updated in real time with a control cycle of 1ms. The limiting protection refers to: setting a threshold for the rate of change of a parameter to eliminate abnormal jumps caused by transient interference from the sensor. For example, the parameter is set as R... s The rate of change threshold is <0.1Ω.

[0075] In step S4, a multi-objective collaborative optimization traction control logic is implemented. This step aims to solve the technical challenge of traditional traction control in achieving high efficiency, high dynamic response, and low noise under complex operating conditions. The system adopts a hierarchical control architecture of "offline map benchmark and online real-time correction, and multi-objective dynamic trade-off" to realize the multi-objective collaborative optimization traction control logic.

[0076] First, initial instructions are generated based on the efficiency map. The controller internally stores a multi-dimensional efficiency map lookup table obtained through offline bench testing. This multi-dimensional efficiency map lookup uses speed ω and torque T as inputs. ref For indexing, the optimal current amplitude I under the corresponding operating condition is pre-stored. opt and phase angle γ opt In other words, in each control cycle, the system first determines the current ω and T... ref Look up the table to obtain the initial current command (i d0 i q0 This step ensures that the motor always operates within its theoretically highest efficiency range under steady-state conditions, avoiding efficiency deviations caused by parameter errors in traditional analytical methods.

[0077] Secondly, a nonlinear magnetic saturation compensation mechanism is integrated: a three-dimensional inductance mapping table considering the coupling effect of direct-axis and quadrature-axis currents is established. In real-time control, accurate inductance values ​​are obtained through online interpolation based on the current current vector coordinates, and the proportional-integral gain of the current loop is dynamically adjusted accordingly to eliminate dynamic performance degradation caused by magnetic saturation. To address the nonlinear inductance changes caused by magnetic saturation of the motor core under heavy loads, the system establishes a three-dimensional inductance mapping table L... d (i d i q ) and L q (i d i q Specifically, this manifests as the following process: real-time calculation based on the current current vector (i... d i q Bilinear interpolation is performed in the three-dimensional table to obtain the accurate dynamic inductance value L. d_real ,L q_real Then, based on the acquired real-time inductance value, the gain parameters of the current loop PI controller are dynamically updated. This gain adaptive technique eliminates the problem of deteriorated dynamic performance of the current loop caused by magnetic saturation, enabling the system to maintain a consistent transient response speed across the entire torque output range.

[0078] Finally, adaptive field weakening and copper loss optimization are implemented. Under low-speed, high-torque conditions, the stator current vector phase angle is fine-tuned based on real-time identified inductance parameters to minimize stator copper losses. Under high-speed, constant-power conditions, the system automatically switches to field weakening control mode, increasing the negative direct-axis current component to offset back electromotive force and expand the speed range. In the low-speed range, based on the initial values ​​from the lookup table, the resistance R identified online in step S3 is used... s and magnetic flux ψ f Real-time correction of phase angle γ further reduces stator copper losses. In the high-speed region of field weakening mode, when the bus voltage approaches its limit, the field weakening algorithm is automatically activated to calculate the required negative direct-axis current i. d_fw To counteract the back electromotive force and ensure error-free tracking of torque commands, the constant power speed regulation range is extended to more than 2.5 times the base speed.

[0079] Furthermore, an integrated intelligent carrier frequency scheduler dynamically adjusts the inverter carrier frequency based on multiple constraints, including speed level and load rate. Under heavy load and low speed conditions, the carrier frequency is increased to suppress torque ripple, while under high speed and stable conditions, the carrier frequency is decreased to reduce switching losses. For example, in high-speed stable cruising conditions with a load rate <30% and a speed >80% of the rated value, the carrier frequency is reduced to the lower limit (e.g., 1kHz), significantly reducing inverter switching losses and improving overall system efficiency. In heavy load and low-speed climbing or rapid acceleration conditions with a load rate >80%, the carrier frequency is increased to the upper limit (e.g., 3-5kHz) to suppress current harmonics and torque ripple, ensuring smooth traction.

[0080] In step S5, a resonance suppression algorithm based on active damping injection is applied to address the electromechanical coupling resonance problem (typical frequency range 50–500 Hz) caused by gearbox clearance, shaft elastic deformation, and distributed capacitance and inductance of long cables in new energy electric locomotives. This step is implemented as follows:

[0081] (1) Extraction of resonant components

[0082] Obtain the rotor speed signal ω from the encoder. m This input is then fed into a bandpass or highpass filter with an adjustable center frequency. The center frequency f of the filter... center It is not a fixed value, but is calculated in real time based on the locomotive's current transmission ratio and wheel diameter, combined with a pre-stored mechanical mode table, to ensure precise locking of the mechanical resonant frequency f. res .

[0083] (2) Adaptive damping generation: The active damping feedback term is generated by extracting the mechanical resonant frequency component from the collected rotor speed signal through bandpass or highpass filtering and then multiplying it by the adaptive damping coefficient.

[0084] The filtered speed fluctuation component Δω resMultiply by the adaptive damping coefficient K d K d The value of K is dynamically adjusted based on the current torque load and vehicle speed: increase K in the low-load, high-resonance-risk region. d In high-load areas, the damping term should be appropriately reduced to prevent current saturation. Its phase is naturally opposite to the direction of mechanical vibration velocity, thus providing negative damping.

[0085] (3) Energy absorption and injection, T damp The torque current setpoint is superimposed on the output of the speed loop. The cutoff frequency of the filter is set to dynamically change with the mechanical resonant frequency of the system, and the phase of the generated damping feedback term is opposite to the direction of the mechanical vibration velocity. The active damping feedback term is injected into the torque current setpoint, and the reverse electromagnetic torque generated by the motor actively absorbs the vibration energy of the mechanical transmission chain and cable system, suppresses the resonance peak of the drive system, and improves the smoothness of locomotive operation.

[0086] In step S6, fault diagnosis and hierarchical fault-tolerant control logic are implemented. To ensure the system's survivability under extreme conditions, this step establishes a hierarchical fault-tolerant system based on multi-source information fusion. The mechanism is as follows: When sensor drift, power device performance degradation, or slight DC bus overvoltage is detected, it is determined to be a level one fault. The system automatically switches to emergency derating mode, which linearly limits the maximum output torque and dynamically reduces the inverter switching frequency to reduce thermal stress and ensure the locomotive travels slowly to the maintenance station. When power device open circuit, severe overcurrent, or severe DC bus overvoltage is detected, it is determined to be a level two fault. This triggers control reconfiguration or a safe shutdown strategy, blocking the faulty bridge arm and switching to non-faulty phases to maintain operation or performing inertial coasting to a stop. Through the above hierarchical response, systemic shutdown accidents are avoided, maximizing the operational safety and availability of the locomotive.

[0087] First, real-time monitoring of multi-dimensional states, including monitoring of thermal state, electrical state, and device health.

[0088] Among them, the thermal state monitoring uses a built-in transient thermal network model based on physical mechanisms to estimate the junction temperature T of IGBT / MOSFET in real time. j The electrical state is monitored by monitoring the DC bus voltage and its rate of change, and the imbalance of the three-phase current is calculated. Device health is monitored by sampling the on-state voltage drop of the power transistor in each switching cycle. If the on-state voltage drop of the sampled power transistor deviates continuously from the initial value, such as an increase of >10%, it is determined that the device performance has degraded, that is, slightly damaged, indicating device aging or poor contact.

[0089] Then, a tiered fault-tolerance decision-making mechanism is implemented. This fault-tolerance decision-making mechanism includes first-level fault tolerance and second-level fault tolerance.

[0090] Level 1 fault tolerance represents degraded operation. The triggering conditions are: slight sensor drift, device performance degradation, and T... j The system is approaching but not exceeding its limit, and the DC bus voltage is slightly overvoltage. The corresponding execution strategy is for the system to enter "limp-home" mode. The controller linearly limits the maximum output torque, such as to 50% of the rated value, and reduces the switching frequency from the normal value to a safe value to reduce switching losses and heat generation. In this mode, the locomotive can still maintain low-speed travel to the nearest maintenance station.

[0091] Level 2 fault tolerance represents reconfiguration or safe shutdown. Triggering conditions are: open / short circuit in the power transistor, severe overvoltage on the DC bus, and severe three-phase current imbalance. The corresponding execution strategy is: immediately block the drive signal of the faulty arm. If it is a single-phase fault, switch to a two-phase operation control algorithm to maintain limited torque output using the remaining two phases; if the fault is too severe to reconfigure, control the locomotive to enter inertial coasting mode, apply electric braking to assist deceleration until a safe stop is reached, and upload a fault code.

[0092] Through this tiered strategy, the system successfully reduced the downtime rate of non-catastrophic failures by 90%. Early warning and de-rated operation can be implemented in the early stages of component aging, avoiding traction converter burnout accidents caused by the escalation of failures, and greatly improving the operational reliability and maintenance economy of locomotives.

[0093] In addition, such as Figure 2 As shown, the present invention also provides a high-efficiency electronic control system for the traction motor of a new energy electric locomotive, used to implement the above-mentioned high-efficiency electronic control method for the traction motor of the new energy electric locomotive, comprising:

[0094] The energy management interface unit is used to connect the power battery and the braking resistor to perform bidirectional energy flow scheduling and bus voltage stabilization control.

[0095] The signal conditioning and acquisition subsystem is used to acquire the three-phase current, bus voltage and temperature signals of the traction motor in real time, and to perform filtering and analog-to-digital conversion on the signals.

[0096] The main control core processing unit communicates with the energy management interface unit and the signal conditioning and acquisition subsystem, respectively, and is used to execute traction control algorithms, online parameter identification, fault diagnosis, and multi-objective optimization decision-making. The main control core processing unit adopts a heterogeneous parallel computing architecture composed of a digital signal processor (DSP) and a field-programmable gate array (FPGA). Data exchange between the DSP and the FPGA is achieved through a high-speed parallel bus. The DSP is configured to execute advanced control logic, including bus voltage feedforward calculation, online motor parameter identification, multi-objective collaborative optimization algorithms, and fault diagnosis logic. The FPGA is configured to handle high-frequency and high real-time tasks, including synchronous trigger sampling of multiple analog signals, hard decoding of encoder signals, generation of space vector pulse width modulation waveforms, and dead time compensation.

[0097] The isolated drive subsystem, connected to the main control core processing unit, is used to convert control signals into drive signals and transmit them with electrical isolation.

[0098] The power conversion module, connected to the isolated drive subsystem, is used to invert DC power into AC power to drive the traction motor.

[0099] The energy feedback management unit is configured to dynamically allocate electric braking power based on the allowable charging power and state of charge of the power battery when the locomotive is braking, and to automatically activate the braking resistor branch when the battery pack cannot absorb all the braking energy, thereby consuming excess energy through pulse width modulation.

[0100] The remaining life prediction module is configured to assess the health status of core components online by recording the historical junction temperature cycle number of the power module, the ripple current load of the capacitor, and the cumulative operating temperature duration of the motor, combined with a preset loss degradation model, and transmit the prediction results to the ground operation and maintenance center.

[0101] Specifically, the signal conditioning and acquisition subsystem is designed with full consideration of the special environment of strong electromagnetic interference inside the electric locomotive. This subsystem integrates multiple synchronous sampling channels, each rigorously configured with a fourth-order Butterworth anti-aliasing filter. Its cutoff frequency is precisely set according to twice the inverter switching frequency to ensure effective filtering of noise pollution generated by high-frequency switching operations. During the analog-to-digital signal conversion, the system uses a 16-bit resolution analog-to-digital converter, coupled with a precision instrumentation amplifier, greatly improving sampling accuracy under small-signal conditions. The current sampling channel employs a differential input method, connected to a through-type low-drift current transformer via shielded twisted-pair cable, utilizing the common-mode rejection characteristics of the differential circuit to cancel induced voltage interference generated by the cable's distributed capacitance.

[0102] The isolated drive subsystem employs a drive chip based on fiber optic communication or magnetic isolation technology, achieving physical-level electrical isolation between the weak current control circuit and the high-voltage, high-power circuit. The drive circuit integrates a current clamping circuit with dynamic latching function, effectively preventing the power transistor from being mis-turned on due to the extremely high voltage change rate generated during the high-speed switching of silicon carbide power devices. Furthermore, the drive circuit also features dynamic Miller clamping and desaturation protection circuitry. By monitoring the voltage drop change of the power transistor in the on-state in real time, once abnormal conditions such as output short circuits are detected, the trigger signal can be forcibly shut off within an extremely short time of 2μs, and a hardware interrupt command can be simultaneously issued to the main control core processing unit to protect the power device from damage. The desaturation protection circuit is configured to monitor the on-state voltage drop of the power transistor in real time. When an overcurrent condition is detected, it will turn off the trigger signal within 2μs and send a hardware interrupt to the main control core processing unit. The isolation drive subsystem also integrates gate voltage active control technology, which is configured to dynamically adjust the gate current by switching through multi-level resistors during the on-state of the switching transistor, and to suppress voltage spikes and reduce turn-off losses by controlling the magnitude of the discharge current during the turn-off state.

[0103] The power conversion module employs high-power-density silicon carbide MOSFETs or IGBTs. An integrated temperature sensor, positioned close to the power chip, captures transient changes in the chip's junction temperature. To ensure thermal stability under high-power operation, the module is mounted on a heatsink with forced air or liquid cooling. Its AC output is equipped with a sine wave filter or dVdt filter composed of a high-voltage inductor and a non-polar film capacitor. The heatsink features a complex fluid flow design, and forced liquid cooling ensures that the module's casing temperature rise is strictly limited to below 40°C at rated output power. The power circuit utilizes a multilayer bus structure. This compact laminated design minimizes parasitic inductance in the bus circuit and suppresses voltage spikes during switching. The DC side of the power conversion module uses a multilayer bus structure and is equipped with high-frequency, long-life film-supported capacitors. The AC output port is equipped with a high-frequency current transformer to monitor common-mode current and cuts off the inverter output when leakage current exceeds a preset threshold.

[0104] The main control core processing unit also includes a historical data storage module, which uses a highly reliable ferroelectric random access memory. The historical data storage module is configured to record all key control parameter waveforms within 1 minute before the locomotive failure occurs, and lock and report the data as black box information when the system experiences an abnormal shutdown. The remaining life prediction module works in conjunction with the historical data storage module to use the fault waveform data recorded by the black box to correct the loss degradation model, so as to achieve full life cycle operating cost optimization and preventive maintenance.

[0105] To verify the technical superiority of the present invention, the following data comparison is conducted with a specific embodiment and a traditional control scheme (comparative example).

[0106] The specific configuration of the embodiment is as follows: a new energy electric locomotive with a rated power of 1200kW is selected, the traction motor is a permanent magnet synchronous motor, the nominal voltage of the DC bus is 1500V, the power module adopts silicon carbide MOSFET, the heat dissipation method is forced liquid cooling, and the switching frequency is adaptively adjusted between 2 and 10kHz.

[0107] The proportional configuration is as follows: a traditional vector control scheme based on fixed parameter PI regulation is adopted, without bus voltage feedforward compensation, without online parameter identification, and the switching frequency is fixed at 4kHz.

[0108] The test conditions were selected as follows: the locomotive towing 3000t of cargo was started on a 20‰ slope, the ambient temperature was set at 35℃, and the bus voltage was simulated to drop by 15% due to the large current discharge of the battery.

[0109] The experimental data are recorded in Table 1:

[0110] Table 1: Comparison of Operating Data between Embodiments and Comparative Examples of the Invention

[0111]

[0112] The data analysis in Table 1 clearly demonstrates that the system and method of this invention exhibit significant engineering advantages across multiple key performance dimensions. Firstly, regarding torque control accuracy, thanks to the combination of the bus voltage feedforward compensation model and high-frequency dynamic sampling technology, the torque ripple in the embodiment under extreme fluctuation conditions is only 4.2%, far lower than the 12.5% ​​of the comparative example. This directly reflects a significant improvement in the smoothness of power output. Secondly, in terms of energy efficiency, the embodiment employs multi-objective collaborative optimization logic and an online parameter identification mechanism, ensuring that the motor always operates at the optimal control point, resulting in a 2.5 percentage point improvement in overall system efficiency. This has significant practical implications for the range of new energy vehicles.

[0113] Finally, the method and system described in this invention fully utilize the parallel computing capabilities of high-performance processors in terms of hardware, and deeply integrate motor control theory with the characteristics of new energy sources in terms of software, constructing a complete traction electric control technology system with high adaptability. This solution is not only applicable to new energy electric locomotives powered by power batteries, but can also be extended to rail transit equipment with various energy sources such as hydrogen fuel cells and supercapacitors, demonstrating strong engineering application value and technological foresight.

[0114] In summary, this invention constructs a complete traction control scheme for new energy electric locomotives through system-level methodological innovation and meticulous engineering design. From the bottom-level signal acquisition resolution and the mid-level parameter adaptive observation to the high-level multi-objective performance management, each level has been deeply optimized for the specific operating conditions of new energy locomotives. Through efficient data interaction and logical collaboration, the various functional modules form a closed-loop control system with strong robustness and global optimality. The implementation of this invention can not only significantly improve the operating efficiency of the traction system and reduce energy consumption, but also effectively enhance the operational stability and safety of the locomotive under complex and changing operating conditions, which has significant technical implications for promoting the green and intelligent transformation of rail transit equipment.

[0115] The specific parameter design and logical implementation of each component and step of this invention strictly follow the latest advancements in power electronic control technology and motor drive theory. In practical engineering deployments, the system can be flexibly configured according to traction motors of different power levels, possessing excellent platform scalability. Through precise control of physical quantities such as current, voltage, and speed, and a profound understanding of the evolution laws of the electromagnetic and temperature fields inside the motor, this invention truly achieves a leap from traditional linear control to modern intelligent and adaptive control, laying a solid technical foundation for the efficient operation of new energy electric locomotives.

[0116] The electronic control system and its control method disclosed in this invention achieve microscopic control of the electromagnetic energy to mechanical energy conversion process through deep coupling of hardware and software. Whether in heavy-load traction in extremely cold regions or high-frequency reciprocating operations in ports and wharves, this system demonstrates excellent control performance. Its strong tolerance to bus voltage fluctuations, real-time correction of motor parameter drift, and refined control of switching losses collectively establish a new technological benchmark in the field of new energy traction electronic control, providing key core technological support for achieving the goal of "carbon peaking and carbon neutrality" in the rail transit sector.

[0117] The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A high-efficiency electronic control method for traction motors in new energy electric locomotives, characterized in that, Includes the following steps: S1. Acquisition of operating status parameters: The operating status parameters of the new energy electric locomotive are acquired in real time through an airborne sensor array; the operating status parameters include DC bus voltage, three-phase stator current of traction motor, rotor mechanical position angle, rotor speed and stator winding temperature; S2. Bus voltage feedforward dynamic compensation: Construct a bus voltage feedforward dynamic compensation model, receive the instantaneous value of DC bus voltage and compare it with the preset nominal bus voltage reference value, and calculate the voltage fluctuation deviation rate. S3. Online identification of key motor parameters: For the stator resistance temperature rise drift and rotor permanent magnet flux thermal decay characteristics of traction motor under heavy load operation, online parameter identification based on multi-physics field coupling characteristics is performed. S4. Multi-objective collaborative optimization traction control: Construct a hierarchical multi-objective collaborative optimization control architecture, based on the maximum torque-current ratio control, and obtain the initial optimal current vector command by combining the pre-stored motor efficiency map online lookup table; S5. Active damping injection resonance suppression: To address the electromechanical coupling resonance problem caused by the complex mechanical transmission chain and long cable connection of new energy electric locomotives, an active damping feedback term is superimposed at the output end of the speed regulator. S6. Fault Diagnosis and Fault-Tolerant Control: Real-time monitoring of key status parameters of the power conversion module, including power device junction temperature, DC bus voltage fluctuation rate, three-phase current imbalance, and on-state voltage drop characteristics of power devices; construction of a multi-level fault diagnosis and fault-tolerant decision-making mechanism.

2. The high-efficiency electronic control method for traction motors in new energy electric locomotives according to claim 1, characterized in that, In step S1, the DC bus voltage is acquired by a Hall closed-loop voltage sensor based on the magnetic compensation principle, and the current loop control gain is dynamically adjusted based on the fluctuation rate of the acquired signal; the three-phase stator current is acquired by a through-hole low-drift current transformer, and the fundamental signal is obtained by synchronous sampling technology synchronized with the extreme point of the pulse width modulation carrier, and then decomposed into excitation and torque components by coordinate transformation; the rotor mechanical position angle and rotor speed are acquired by an absolute value photoelectric encoder with differential signal output function, and a sensorless control backup scheme based on sliding mode observer is configured to automatically switch to estimation mode when the main encoder fails; the stator winding temperature is monitored in real time by a thermal resistance sensor embedded in the stator winding.

3. The high-efficiency electronic control method for traction motors in new energy electric locomotives according to claim 1, characterized in that, In step S2, the voltage fluctuation deviation rate is introduced into the modulation index correction link of the space vector pulse width modulator to adjust the conduction time of the inverter switching transistor in real time, so as to offset the influence of bus voltage fluctuation on the equivalent output voltage vector. The compensation model is equipped with a voltage safety window and amplitude limiting protection logic. When the voltage fluctuation exceeds the safety window threshold, the correction amplitude of the modulation index is automatically limited to prevent the current waveform distortion caused by entering the deep overmodulation region, thereby maintaining the excitation current and torque current at the stator end of the motor in a decoupled state under voltage fluctuation environment. When processing the voltage signal, the bus voltage feedforward dynamic compensation model uses a sliding window weighted average filtering algorithm to remove non-characteristic voltage spikes caused by unstable pantograph current collection or internal impedance switching of the power battery.

4. The high-efficiency electronic control method for traction motors in new energy electric locomotives according to claim 1, characterized in that, The specific operation process of step S3 is as follows: Collect the stator winding temperature signal and the current loop residual signal, construct an adaptive observer, and dynamically update the stator resistance and rotor flux parameters in the prediction model by comparing the current estimation value output by the motor prediction model with the measured current value and using the error feedback mechanism; The online parameter identification adopts the working condition adaptive time-division identification logic: during the locomotive steady-state cruise phase, lock the rotor flux parameters and prioritize the online identification of stator resistance; during the high-current dynamic switching phase such as locomotive acceleration or braking, lock the stator resistance parameters and prioritize the online calibration of rotor flux. Furthermore, the adaptive observer also integrates a high-frequency signal injection auxiliary mechanism: when the motor is running at low speed, a high-frequency disturbance signal is injected into the stator winding, and the rotor flux linkage and inductance parameters are extracted based on the motor's impedance response to the signal; the real-time updated stator resistance and rotor flux linkage parameters are smoothed and filtered before being fed back to the current loop controller to eliminate torque control deviations caused by parameter mismatch; wherein, the online parameter identification adopts a time-division identification strategy: during the locomotive's steady-state cruise phase, online identification of the stator resistance is prioritized; during the high-current dynamic switching phase of the locomotive's acceleration or braking, online calibration of the rotor flux linkage is prioritized. The identification process also introduces an identification method based on injected signals: during motor operation, a high-frequency non-characteristic current vector is injected into the stator current, the frequency of which is higher than the motor's mechanical response frequency.

5. The high-efficiency electronic control method for traction motors in new energy electric locomotives according to claim 1, characterized in that, In step S4, the hierarchical multi-objective collaborative optimization control architecture is generated based on the initial instructions of the efficiency map; it integrates a nonlinear magnetic saturation compensation mechanism: a three-dimensional inductance mapping table considering the coupling effect of direct-axis and quadrature-axis currents is established, and the accurate inductance value is obtained by online interpolation based on the current current vector coordinates in real-time control, and the proportional-integral gain of the current loop is dynamically adjusted accordingly to eliminate the dynamic performance degradation caused by magnetic saturation; adaptive field weakening and copper loss optimization: under low-speed, high-torque conditions, the stator current vector phase angle is finely adjusted based on the real-time identified inductance parameters to minimize stator copper losses; under high-speed, constant-power conditions, it automatically switches to field weakening control mode, and expands the speed range by increasing the negative direct-axis current component to offset the back electromotive force; the inverter carrier frequency is dynamically adjusted according to the constraints of speed level and load rate; the carrier frequency is increased under heavy-load, low-speed conditions to suppress torque ripple, and the carrier frequency is decreased under high-speed, stable conditions to reduce switching losses.

6. The high-efficiency electronic control method for traction motors in new energy electric locomotives according to claim 1, characterized in that, In step S5, the active damping feedback term is generated by extracting the mechanical resonant frequency component from the acquired rotor speed signal through bandpass or high-pass filtering, and then multiplying it by an adaptive damping coefficient. The cutoff frequency of the filter is set to dynamically change with the mechanical resonant frequency of the system, and the phase of the generated damping feedback term is opposite to the direction of the mechanical vibration velocity. The active damping feedback term is injected into the torque current setpoint, and the reverse electromagnetic torque generated by the motor actively absorbs the vibration energy of the mechanical transmission chain and cable system, suppresses the resonance peak of the drive system, and improves the smoothness of locomotive operation.

7. The high-efficiency electronic control method for traction motors in new energy electric locomotives according to claim 1, characterized in that, In step S6, the multi-level fault diagnosis and fault-tolerant decision-making mechanism is as follows: when sensor drift, power device performance degradation, or slight DC bus overvoltage is detected, it is determined to be a level one fault, and the system automatically switches to emergency derating operation mode. By linearly limiting the maximum output torque and dynamically reducing the inverter switching frequency, thermal stress is reduced and the locomotive is ensured to travel slowly to the maintenance station. When power device open circuit, severe overcurrent, or severe DC bus overvoltage is detected, it is determined to be a level two fault, and the control reconfiguration or safety shutdown strategy is triggered. The faulty bridge arm is blocked and the system switches to the non-faulty phase to maintain operation or performs inertial coasting stop.

8. A high-efficiency electronic control system for a traction motor in a new energy electric locomotive, used to implement the method described in any one of claims 1 to 7, characterized in that, include: The energy management interface unit is used to connect the power battery and the braking resistor to perform bidirectional energy flow scheduling and bus voltage stabilization control. The signal conditioning and acquisition subsystem is used to acquire the three-phase current, bus voltage and temperature signals of the traction motor in real time, and to perform filtering and analog-to-digital conversion on the signals. The main control core processing unit is communicatively connected to the energy management interface unit and the signal conditioning and acquisition subsystem, respectively, and is used to execute traction control algorithms, online parameter identification, fault diagnosis, and multi-objective optimization decision-making. The main control core processing unit adopts a heterogeneous parallel computing architecture composed of a digital signal processor (DSP) and a field-programmable gate array (FPGA). The DSP and FPGA exchange data through a high-speed parallel bus. The DSP is configured to execute advanced control logic, including bus voltage feedforward calculation, online motor parameter identification, multi-objective collaborative optimization algorithms, and fault diagnosis logic. The FPGA is configured to handle high-frequency and high real-time tasks, including synchronous trigger sampling of multiple analog signals, hard decoding of encoder signals, generation of space vector pulse width modulation waveforms, and dead time compensation. An isolated drive subsystem, connected to the main control core processing unit, is used to convert control signals into drive signals and transmit them with electrical isolation. A power conversion module, connected to the isolated drive subsystem, is used to invert DC power into AC power to drive the traction motor. The energy feedback management unit is configured to dynamically allocate electric braking power according to the allowable charging power and state of charge of the power battery when the locomotive is braking, and automatically activate the braking resistor branch when the battery pack cannot absorb all the braking energy to consume excess energy through pulse width modulation. The remaining life prediction module is configured to assess the health status of core components online by recording the historical junction temperature cycle number of the power module, the ripple current load of the capacitor, and the cumulative operating temperature duration of the motor, combined with a preset loss degradation model, and transmit the prediction results to the ground operation and maintenance center.

9. A high-efficiency electronic control system for traction motors in new energy electric locomotives according to claim 8, characterized in that, The signal conditioning and acquisition subsystem includes multiple synchronous sampling channels, each of which is sequentially equipped with an anti-aliasing filter, a precision instrumentation amplifier, and an analog-to-digital converter. The anti-aliasing filter adopts a fourth-order Butterworth structure, and its cutoff frequency is set according to twice the switching frequency of the power conversion module. The current sampling channel adopts a differential input method and is connected to the current transformer through shielded twisted-pair cable. The isolated drive subsystem employs a drive chip based on fiber optic communication or magnetic isolation technology, featuring dynamic Miller clamping and desaturation protection circuitry. The desaturation protection circuitry is configured to monitor the on-state voltage drop of the power transistor in real time, and upon detecting an overcurrent condition, to shut down the trigger signal within 2μs and issue a hardware interrupt to the main control core processing unit. The isolated drive subsystem also integrates active gate voltage control technology, configured to dynamically adjust the gate current during the transistor's on-state phase via multi-stage resistor switching, and to suppress voltage spikes and reduce turn-off losses during the turn-off phase by controlling the discharge current magnitude. The power conversion module uses high-power-density silicon carbide MOSFETs or IGBTs, and integrates a temperature sensor closely attached to the power chip. The power conversion module is mounted on a heat sink with a forced air cooling or liquid cooling structure, and its AC output terminal is equipped with a sine wave filter or dVdt filter composed of a high-voltage inductor and a non-polar thin-film capacitor. The DC side of the power conversion module adopts a stacked bus structure and is equipped with a high-frequency, long-life thin-film supporting capacitor. The AC output port is equipped with a high-frequency current transformer to monitor the common-mode current and cut off the inverter output when the leakage current exceeds a preset threshold.

10. A high-efficiency electronic control system for traction motors in new energy electric locomotives according to claim 8, characterized in that, The main control core processing unit also includes a historical data storage module, which uses a highly reliable ferroelectric random access memory. The historical data storage module is configured to record all key control parameter waveforms within one minute before the locomotive failure occurs, and lock and report the data as black box information when the system experiences an abnormal shutdown. The remaining life prediction module works in conjunction with the historical data storage module to use the fault waveform data recorded by the black box to correct the loss degradation model, so as to achieve full life cycle operating cost optimization and preventive maintenance.

Images