A method and system for optimizing motor control for a vehicle

Abstract

The present invention discloses a method and system for optimizing motor control for a vehicle. First, a data acquisition module collects vehicle data of the entire vehicle, and dynamically analyzes the real-time acquisition of the vehicle data to obtain the complexity of the vehicle data. Combining the complexity of the data acquisition state, it analyzes the boundary threshold of the change of data related to motor control. Then, it analyzes the motor parameters through the vehicle data to obtain an interference evaluation boundary, and combines the boundary threshold to analyze the adjustable range boundary of the direct control parameters of the motor. The dynamic analysis of the data acquisition state of the vehicle data increases the accuracy of the analysis of motor-related data, making the range of motor control optimization more reasonable. Finally, an adaptive optimization control of the motor control is performed, greatly improving the accuracy of the adaptive control optimization algorithm in analyzing the changes in the vehicle operating environment, and making the motor control system obtain more precise control optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle control, and particularly relates to a method and system for optimizing motor control for a vehicle. Background Art

[0002] A vehicle motor controller is the nerve center connecting the motor and the battery, and is used to adjust various performances of the vehicle. By receiving instructions from the driver, it can achieve precise control of parameters such as the rotational speed, torque, and direction of the motor.

[0003] As the load and environmental conditions continuously change during the operation of the vehicle, the motor control optimization algorithm processes a large amount of data, makes real-time adjustments and optimal control of the motor parameters, and makes quick decisions. For example, in the prior art, the adaptive control algorithm improves the motor performance by improving the vehicle control algorithm and system design, and can adjust the motor parameters in real time to adapt to complex changes. However, the control accuracy of the adaptive control algorithm is affected by the quality and acquisition status of vehicle data collection. For example, the amount of vehicle data collected in real time and data errors. When the battery power is insufficient and the road conditions are poor, the sensor acquisition fails, resulting in problems such as a huge change in the amount of vehicle data and an increase in errors. And the adaptive control optimization algorithm is based on the low-quality vehicle data of the faulty sensor, which leads to misjudgment and over-optimization of the optimization algorithm. In order to improve the optimization accuracy and system stability of the adaptive control optimization algorithm, and enable the vehicle to correspond to the corresponding control optimization degree under different complex environmental levels, the present invention proposes a method and system for optimizing motor control for a vehicle. Summary of the Invention

[0004] In order to solve the above problems, that is, to solve the problems mentioned in the background art, the present invention provides a method and system for optimizing motor control for a vehicle. First, dynamically analyze the acquisition situation of vehicle data to obtain the real-time dynamic increase in the complexity of vehicle data. Combine the analysis of the complexity of the data acquisition status to analyze the boundary threshold between motor control and data changes. Analyze the motor parameters through vehicle data, and then perform error analysis based on vehicle data to obtain the interference evaluation boundary. Use the interference evaluation boundary to represent the correction of vehicle data acquisition errors, and combine the boundary threshold to analyze the adjustable range of the direct control parameters of the motor. By gradually analyzing, the fitting degree of motor data is increased, making the range of motor control optimization more reasonable. When the vehicle data changes complexly, use the real-time increase in the complexity of vehicle data, the boundary threshold, and the adjustable boundary to reduce the dimension of the data, and finally perform adaptive optimization control of the motor control, greatly improving the accuracy of the adaptive control optimization algorithm for analyzing changes in the vehicle operating environment, and enabling more precise control optimization of the motor control system.

[0005] The present invention proposes a method for optimizing motor control for a vehicle, which optimizes the control parameters of the motor controller according to the vehicle data collected by sensors. The specific analysis process is as follows:

[0006] Step 1: The vehicle data during the vehicle driving process includes all data generated by the entire vehicle. The real-time data of the motor operation is obtained through sensors, and the real-time state of the total vehicle data collection is analyzed to obtain the complexity corresponding to the real-time state change at the i-th moment ;

[0007] Step 2: The complexity of vehicle data collection under different adaptive scenarios corresponding to different complex environments is different. The motor parameters that actually affect the motor operation control in the vehicle data include direct control parameters and indirect parameters. All motor parameters in the vehicle data of the motor operation at different moments are recorded as a multi-dimensional state vector, and the boundary threshold of motor control is obtained through multi-dimensional state vector analysis;

[0008] Step 3: Analyze the error interference of the vehicle data collected by the sensors during the motor operation to obtain the interference evaluation boundary of data collection, and then calculate the adjustable range boundary of the dynamic drive of the directly adjustable range of the motor in combination with different adaptive scenarios;

[0009] Step 4: Determine the range of particle search through the adjustable range boundary of motor drive and the interference evaluation boundary of data collection, and optimize the parameters through the particle swarm search algorithm, and transform the non-linear relationship

[0010] and determine the optimal parameters of the system control algorithm through the optimal position searched;

[0011] Step 5: Use the neural network algorithm to optimize the control parameters of the motor, construct the dynamic models of the vehicle system and each component of the system, and combine the vehicle speed, engine speed, state of charge of the battery, SOC vehicle state signal and the neural network to predict the vehicle speed and torque.

[0012] Specifically, the vehicle data of the motor is the operation collection data of the entire vehicle, which is recorded as historical operation data and real-time data from the time dimension. Extract the detection indexes corresponding to all change parameters in the motor operation adaptive scenario, and record them as represents the number of detection indexes that change with the adaptive scenario in the vehicle data, i represents the extraction moment in the time dimension, and the real-time state of the vehicle data collected at different moments is analyzed. The total number of detection indexes in the vehicle data is N, and the number of changes in the detection indexes corresponding to different adaptive scenarios at different moments is recorded as Then, analyze the change of the detection indexes corresponding to each detection moment i to obtain the index change ratio of the real-time change of data collection , and then calculate the complexity of the real-time change of vehicle data corresponding to different motor adaptive scenarios according to the index change , calculate the ratio of the change of the detection index data , , , where , and respectively represent the change values of the detection index in the adaptive scenario at the i-th moment and the detection index at the (i - 1)-th moment, .

[0013] Specifically, in the second step, the number of parameters of the motor parameters changes dynamically under different adaptive scenarios, and the corresponding adjustment range changes accordingly. Dynamically analyze the characteristics of the dynamic change boundary of the data related to the motor according to the historical operation data in the vehicle data to obtain the boundary threshold of the adaptive scenario. Denote the change of the detection index corresponding to the i-th moment and the number of indirect parameters related to the motor as , and denote the number of changes in the detection index of the direct control parameters of the motor as , the complexity is . Analyze a point set according to the control equation of the boundary. The point set includes the control data of the motor parameters. Then establish an elliptic partial differential equation of the dynamic boundary based on the point set, and analyze the extreme value of the elliptic differential equation with the point set as the value range. Then calculate the boundary threshold according to the extreme value , the equation is as follows:

[0014]

[0015] Among them, is the elliptic function of the dynamic boundary feature, is the auxiliary variable, that is, the complexity of the data change at the i-th moment, is the time change function obtained according to the historical data, is the change function of the corresponding feature of the dynamic boundary, and are obtained according to the historical data of the motor operation. Then, by solving the extreme value of the elliptic partial differential equation and using the extreme value for analysis, the boundary threshold is obtained.

[0016] Specifically, in the third step, the error evaluation of the vehicle data collected by the sensor is used to obtain the interference evaluation boundary. In an adaptive scenario, the sensor collects the operation data of the vehicle. The transmission signals of different sensors affect each other in the same adaptive scenario. Perform hierarchical analysis on the vehicle data collected by all sensors. In the th layer the probability that , then based on the binomial distribution, error analysis is performed on the sensor data, the confidence interval of the data is calculated, the confidence level of the vehicle data is determined, and finally the interference evaluation boundary of the error is calculated. .

[0017] Furthermore, as the vehicle operating environment and motor operating parameters change, the most suitable control parameters of the motor change. Calculate the adjustable range boundary of the dynamic drive corresponding to different adaptive scenarios of motor operation. Through the training of historical operation data by the neural network algorithm, the control range of the direct control parameters of motor operation is obtained. The input vector of the system is The output is a q-dimensional vector , and at the same time and The relationship is , determine to use the neural function to approximate the boundary threshold , is the corresponding point value of the interference evaluation boundary, is the data change dimension in the adaptive scenario, and the adjustable range boundary is , satisfying the following performance index function:

[0018] .

[0019] Furthermore, within the adjustable range of the motor, use the control strategy of the particle swarm algorithm in the neural network system to further optimize the direct control parameters of the motor. The particle swarm is described as , there are particles in the multi-dimensional target search space. Determine the position of the particles in the search space. After calculating the objective function, the individual extreme value and the group extreme value of the th particle are obtained, and are denoted as and respectively, , . According to the principle that after each update of each particle, its own fitness value is compared with the individual extreme value and the group extreme value, so as to continuously update to determine the optimal fitness positions of the group and the individual.

[0020] The direct control parameters of the motor include speed and torque. During the optimization process of the motor control parameters, dynamic analysis of the vehicle data of the whole vehicle is carried out, and then the control parameters of the motor are optimized based on the particle swarm algorithm in the neural network algorithm.

[0021] The present invention also provides an optimized motor control system for a vehicle. The system includes a control module, a data acquisition module, a boundary analysis module, a parameter optimization module, and a dynamic adaptation module. The control module is the core of the algorithm, which realizes intelligent control of the motor by sending different control signals to the motor of the vehicle and coordinates the joint control of each module in the vehicle. The data acquisition module includes a speed sensor, a pressure sensor, a humidity sensor, and a temperature sensor, which collect the current signals of the motor under different adaptive scenarios and then convert different signals into the form of vehicle data for storage. The boundary analysis module, the dynamic adaptation module, and the parameter optimization module optimize and analyze the data of the motor operation.

[0022] The beneficial effects of the present invention are as follows:

[0023] 1. In the present invention, the neural network algorithm is used to intelligently control the motor parameters, so that the motor adapts to the environment during operation, and the vehicle motor can dynamically adjust the control parameters of the motor according to the real-time state and driving environment to achieve more efficient and stable power output. When operating in a complex environment, it can make each subsystem operate in coordination, thereby improving the efficiency and stability of the system.

[0024] 2. In the present invention, the vehicle data of the entire vehicle during motor operation is analyzed. When the vehicle load and the motor operation environment change, the optimal control parameters of the motor also change. Although the control parameters can be optimized by intelligent algorithms in the prior art, the analysis algorithm is restricted by the data quality. When the operation environment changes complexly, the amount of vehicle data increases, and the adaptive optimization algorithm needs to process a large amount of data and perform complex calculations to optimize the parameters, resulting in a more obvious delay in the system, which affects the power performance and response speed of the vehicle. In the present invention, the complexity of data change is obtained by analyzing the real-time state of the acquisition and update of the vehicle data of the entire vehicle. When the vehicle operation scenario is complex, the number of real-time updated data items of the vehicle increases, and the complexity of data change increases. Then, combined with the analysis of the complexity of the data acquisition state, the boundary threshold of the motor control and data change is analyzed. According to the parameter changes related to the motor in the vehicle data, the change in the entire vehicle data is analyzed to obtain the boundary threshold of the motor in different adaptive scenarios. In this process, not only the real-time performance of the data in the adaptive process is improved, but also the efficiency of the motor intelligent control is improved.

[0025] 3. The complexity and boundary threshold of the data are obtained through the acquisition and monitoring of the vehicle's overall vehicle data. Then, the motor parameters are analyzed based on the vehicle data, and the interference evaluation boundary is obtained through error analysis of the vehicle data. The interference evaluation boundary is used to represent the correction of the vehicle data acquisition error. The adjustable range of the direct control parameters of the motor is analyzed in combination with the boundary threshold. By analyzing step by step, the fitting degree of the motor data is increased, making the optimization range of the motor control more reasonable. When the vehicle data changes complexly, the dimensionality reduction of the data is performed using the complexity, boundary threshold, and adjustable boundary that increase in real time with the vehicle data. Finally, the adaptive optimization control of the motor control is performed, greatly improving the accuracy of the adaptive control optimization algorithm for analyzing changes in the vehicle operating environment, and making the motor control system obtain more precise control optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is the analysis flow chart of the present invention;

[0027] Figure 2 is the overall module diagram in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] The following refers to the attached Figure 1 to the attached Figure 2 to describe the preferred embodiments of the present invention. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention. Specific embodiments:

[0029] With the continuous development of new energy vehicle technology, motor control technology is a key core technology, which is of great significance for improving vehicle performance and realizing intelligent driving. Therefore, improving the motor control and optimization control decision-making of new energy vehicles is of great significance for motor control; the drive motor controller is mainly composed of a control main board, drive, interface circuit, capacitor, resistor, sensor, etc., and needs to provide functions such as idling, motor forward rotation, energy recovery, slope parking, communication protection, etc. The motor controller responds to the control and sends feedback signals to realize the control and adjustment of the running state of the drive motor; the intelligence and adaptability of motor control are also important directions for future development. When the vehicle load and the motor operating environment change, the optimal control parameters of the motor also change accordingly. The control parameters are optimized through an adaptive analysis algorithm. However, the adaptive analysis algorithm is restricted by the data quality. When the operating environment changes complexly, the amount of vehicle data increases, and the adaptive optimization algorithm needs to process a large amount of data and perform complex calculations to achieve optimized parameters, resulting in a more obvious delay in the system, affecting the power performance and response speed of the vehicle. In order to improve the efficiency of vehicle motor control, the present invention proposes a motor control optimization method for vehicles, which optimizes the control parameters of the motor controller according to the vehicle data collected by the sensor. The specific analysis process is as follows:

[0030] Step 1: The vehicle data during the vehicle driving process includes all data generated by the whole vehicle. The real-time data of the motor operation is obtained through sensors, and the real-time state of the total vehicle data acquisition is analyzed to obtain the complexity corresponding to the real-time state change at the i-th moment. ; As the vehicle operation environment changes, the real-time acquisition and update status of the vehicle data are constantly changing. In a complex environment, the number of detection items that change in the vehicle data increases, and the more complex the real-time state change is, the more it is utilized.

[0031] Step 2: The complexity of the vehicle data during vehicle operation in different complex environment scenarios is different, and the number of data indirectly related to the vehicle motor operation changes with the adaptive environment. In order to improve the efficiency of adaptive control and reduce the complex data analysis process, the boundary thresholds of the vehicle data changes in different adaptive scenarios are analyzed. The motor parameters that actually affect the motor operation control in the vehicle data include direct control parameters and indirect parameters. All motor parameters in the vehicle data related to the motor operation at different moments are recorded as a multi-dimensional state vector, and the boundary threshold of motor control is obtained through multi-dimensional state vector analysis.

[0032] Step 3: The error interference analysis of the vehicle data collected by the sensors during the motor operation is carried out to obtain the interference evaluation boundary of the data acquisition, and then the adjustable range boundary of the dynamic drive of the directly adjustable range of the motor is calculated in combination with different adaptive scenarios. The control range of the control parameters of the motor is within the range defined by the adjustable boundary.

[0033] Step 4: The range of particle search is determined through the adjustable range boundary of the motor drive and the interference evaluation boundary of the data acquisition, and the parameters are optimized through the particle swarm search algorithm to transform the non-linear relationship

[0034] and the optimal parameters of the system control algorithm are determined through the optimal position searched.

[0035] Step 5: The neural network algorithm is used to optimize the control parameters of the motor, a dynamic model of the vehicle system and each component of the system is constructed, and the vehicle speed, engine speed, battery state of charge, SOC vehicle state signal, and neural network are combined to predict the future vehicle speed and torque.

[0036] Common motor control algorithms in the prior art also include vector control, direct torque control, field-oriented control, pulse width modulation, sliding mode control, etc. Optimization methods such as adaptive control algorithms and vector control algorithms improve the motor performance by improving the vehicle control algorithm and system design.

[0037] Motor controllers usually consist of components such as a control module, a driver, and a power conversion module. These components work together to convert the direct current of the battery into the alternating current required by the motor and achieve precise control of the motor. The parameter optimization of motor control algorithms is a complex systematic process that requires comprehensive consideration of multiple factors such as the characteristics of the motor, the application scenario, and the control requirements.

[0038] Neural network control: Utilize the neural network to learn and adapt to the nonlinear characteristics of the system. By training the neural network, it can output optimal control parameters.

[0039] Adaptive control: Automatically adjust the control parameters according to the real-time operating state of the system and changes in the external environment. Adaptive control algorithms can cope with various uncertainties and disturbances, improving the robustness and stability of the system.

[0040] In electric vehicles, the choice of motor control method depends on the specific requirements and application scenarios of the vehicle. For example, in situations where high performance and high efficiency are required, vector control or direct torque control may be selected; while in situations with strict cost requirements, more economical control methods such as voltage control or current control may be chosen.

[0041] The vehicle data of the motor collects data for the operation of the entire vehicle. Recorded from the time dimension as historical operation data and real-time status data, extract the detection indicators corresponding to all changing parameters in the motor operation adaptive scenario, denoted as represents the number of detection indicators that change with the adaptive scenario in the vehicle data, i represents the extraction moment in the time dimension. Conduct real-time status analysis on the vehicle data collected at different moments. The total number of detection indicators in the vehicle data is N, and the number of changes in the detection indicators corresponding to different adaptive scenarios at different moments is denoted as Then analyze the change in the detection indicators corresponding to each detection moment i to obtain the index change ratio of the real-time change in data collection Then calculate the complexity of the real-time change in the vehicle data corresponding to different motor adaptive scenarios based on the index change Calculate the ratio of the change in the detection indicator data where and respectively represent the change values of the same detection indicator under the adaptive scenario at the i-th moment and the detection indicator at the (i - 1)-th moment. Different operating environments correspond to different adaptive scenarios. Use complexity to represent the change in the data collection state brought about by the environment.

[0042] ​​​In the second step, the number of parameters of the motor parameters varies dynamically under different adaptive scenarios, and the corresponding adjustment range changes accordingly. The characteristics of the dynamic boundary of the adjustment range change of the motor are dynamically analyzed based on historical operation data to obtain the boundary threshold of the adaptive scenario. The change in the detection index corresponding to the current i-th moment and the number of indirect parameters related to the motor are denoted as ; the change in the detection index of the direct control parameter of the motor is denoted as ; the complexity is . A point set is obtained through the analysis of the control equation of the boundary. The point set includes the control data of the motor parameters. Then, an elliptic partial differential equation of the dynamic boundary is established based on the point set, and the extreme value of the elliptic differential equation is analyzed with the point set as the value range. Then, the boundary threshold is calculated based on the extreme value . The equation is as follows:

[0043]

[0044] Among them, is the elliptic function of the dynamic boundary characteristics, is the auxiliary variable, that is, the complexity of the data change at the i-th moment, is the time change function obtained from historical data, is the change function of the characteristics corresponding to the dynamic boundary, and are obtained from the historical data of the motor operation. Then, by solving the extreme value of the elliptic partial differential equation and using the extreme value for analysis, the boundary threshold is obtained. The boundary threshold represents the boundary of the data change related to the motor control in the vehicle data change of the whole vehicle.

[0045] The control processes of different types of motors are different. For example, for permanent magnet synchronous motors and three-phase motors, there are four parameters for the electrical characteristics of permanent magnet synchronous motors, namely the stator resistance of the motor, d-axis inductance, q-axis inductance, and permanent magnet flux linkage. For the sake of simple calculation, most control methods usually assume that these parameters are fixed. However, due to manufacturing errors and the influence of temperature and stator current saturation during the operation of permanent magnet synchronous motors, the parameters of permanent magnet synchronous motors will change accordingly and are difficult to measure in real time. The original control methods cannot achieve good control effects.

[0046] In the third step, the interference evaluation boundary is obtained by evaluating the error of the vehicle data collected by the sensor. In an adaptive scenario, the sensor collects the operation data of the vehicle. The transmission signals of different sensors affect each other in the same adaptive scenario. The vehicle data collected by all sensors is analyzed hierarchically. In the th layer, The probability that a sensor is assigned to the same group follows a binomial distribution. , then based on the binomial distribution, error analysis is performed on the sensor data, the confidence interval of the data is calculated, the confidence level of the vehicle data is determined, and finally the interference evaluation boundary of the error is calculated .

[0047] As the environment in which the vehicle operates and the operating parameters of the motor change, the most suitable control parameters of the motor change. Calculate the adjustable range boundary of the dynamic drive corresponding to different adaptive scenarios of the motor operation. Through the training of the neural network algorithm on the historical operation data, the control range of the direct control parameters of the motor operation is obtained. The input vector of the system is , and the output is a q-dimensional vector , and at the same time and The relationship is , determine to use the neural function to approximate the boundary threshold , is the corresponding point value of the interference evaluation boundary, is the data change dimension in the adaptive scenario, and the adjustable range boundary is , satisfying the following performance index function:

[0048] .

[0049] Through the relationship between the dynamic response of the system to control the motor and the mapping, the mapping relationship between the input and output of the system, the non-linear relationship and the linear relationship, the state equation describing the input-output relationship of the system:

[0050]

[0051] By setting an operator ,

[0052] Specifically described as as the input of the multivariable control, is the intermediate control quantity. Assume the system input dimensional vector , and at the same time the output is dimensional vector , and at the same time and The relationship is: , where is defined as a set of vectors that have a functional relationship at a certain moment and are continuous in the domain. Specifically described as , where , it is known that of The nth derivative is , if the operator satisfies the following formula

[0053]

[0054] Thus, the order inverse system of the original system can be obtained. When the system is invertible, the inverse system can be considered as a realization method, characterizing the length of the pure integral link in the original system.

[0055] is the n-dimensional state vector of the system,

[0056] is the dimensional input vector of the system,

[0057] is the dimensional output vector of the system,

[0058] and have a locally analytic multivariable nonlinear mapping relationship.

[0059] Within the adjustable range of the motor , the direct control parameters of the motor are further optimized using the control strategy of the particle swarm algorithm in the neural network system. The particle population is described as , and there are particles in the multi-dimensional target search space. The positions of the particles are determined within the search space. After calculating the objective function, the individual extreme value and the global extreme value of the th particle are obtained, denoted as and respectively, , . According to the principle that after each update of each particle, its fitness value is compared with the individual extreme value and the global extreme value, and then updated continuously to determine the optimal fitness positions of the population and the individual.

[0060] For example, the speed and the position of the motor are updated to find the speed corresponding to the most suitable control. At the same time, , are respectively restricted between the positive and negative maximum speed and position values. The formulas for updating the speed and position can be obtained: ,

[0061] where is the inertia weight, is the th iteration, is the velocity of the particle at the -th iteration, and are the acceleration factors; and The random numbers are set to be between .

[0062] The intelligent control algorithm of the vehicle motor also includes the PID control algorithm, fuzzy control algorithm, neural network control algorithm, etc. The mathematical expression formula of the PID controller:

[0063]

[0064] Among them, is the proportional coefficient, is the integral time constant, is the differential time constant, is the difference between the given input signal and the actual output signal,

[0065]

[0066] Among them, in the fitness function , are respectively:

[0067]

[0068] Among them, T is the time of the simulation run, is the error between the given value and the output value of the system. Among them , , are the weight coefficients, is the system regulation time, is the system rise time, is the number of maximum values of the rotational speed during the operation of the motor, is the maximum value of the rotational speed, is the target rotational speed.

[0069] The direct control parameters of the motor include speed and torque. During the optimization process of the motor control parameters, the vehicle data of the whole vehicle is dynamically analyzed, and then the control parameters of the motor are optimized based on the particle swarm algorithm in the neural network algorithm.

[0070] The optimization process of the motor:

[0071] 1. Set the target to be optimized and clarify the optimization goal;

[0072] 2. Establish a motor model, establish the mathematical model of the motor, which can describe the characteristics and behaviors of the motor, including the relationship between the input voltage, current and the output rotational speed, torque, and further analyze the control strategies and algorithms of the motor;

[0073] 3. Select optimization algorithms: PID control algorithm, fuzzy control algorithm, neural network control algorithm;

[0074] 4. Evaluate and optimize the performance. The evaluation metrics may include speed control accuracy, steady-state error, dynamic response time, and energy consumption. Further optimize and adjust according to the evaluation results.

[0075] An optimized motor control system for a vehicle. The system includes a control module, a data acquisition module, a boundary analysis module, a parameter optimization module, and a dynamic adaptive module. The control module is the core of the algorithm, which realizes intelligent motor control by sending different control signals to the vehicle's motor and coordinates the joint control of each module in the vehicle. The subroutine of the data acquisition module collects information such as the motor speed and rotor position of the motor, providing data support for the control of the driving motor operation state. The status detection subroutine judges whether a fault occurs during the operation of the system; The data acquisition module includes a speed sensor, a pressure sensor, a humidity sensor, and a temperature sensor. A sensor is a device used to detect, measure, and sense environmental physical quantities, converting physical phenomena into measurable and perceivable electrical signals and transmitting them to a control system or other devices for processing. Sensors include various types, including optical sensors, temperature sensors, pressure sensors, and sound sensors. Each sensor has its detection range, accuracy, and response;

[0076] Collect the current signals of the motor under different adaptive scenarios, and then convert different signals into the form of vehicle data for storage. The boundary analysis module, dynamic adaptive module, and parameter optimization module optimize and analyze the data of the motor operation. In order to achieve the optimization of the acceleration performance, adjust the parameters and optimize the algorithm of the motor control system; The key parameters include the settings of the current, voltage, and torque of the motor controller, as well as the PID parameters in the motor control algorithm. By reasonably adjusting these parameters, the motor outputs greater power and torque during the acceleration process, thereby improving the acceleration performance.

[0077] The participation of the battery management system BMS also needs to be considered for the optimization of the acceleration performance. Dynamically adjust the output power of the battery according to the state and performance characteristics of the battery;

[0078] The field excitation control strategy of the motor controller can also be adopted. By adjusting the field excitation parameters of the motor control, the magnetic field distribution is changed, thereby increasing the torque output of the motor; The motor control system should be matched with the vehicle's transmission system, drive system, etc. to ensure that the power output of the motor can be fully exerted. Measures such as optimizing the vehicle's mass distribution and reducing air resistance can also be taken to improve the acceleration performance of the entire vehicle.

[0079] During the intelligent control process of the motor, the three-phase currents output by the inverter circuit detected by the current sensor , , , after being converted from the three-phase stationary coordinate system to the two-phase stationary coordinate system, the stator current information , and the stator flux linkage information output by the stator flux linkage observer , are simultaneously used as the input of the torque observer, and the output electromagnetic torque information and the torque coefficient information .

[0080]

[0081] The difference between the reference torque value and the observed torque value is used as the output of the torque regulator, and the output logical switch quantity ,

[0082]

[0083] The relationship between torque and stator and rotor flux linkages is

[0084]

[0085] When and are constant in magnitude, is proportional to , and approximately proportional to , so the action of the space voltage vector can be adjusted to control , and further control to change.

[0086] When the present invention is specifically used, the vehicle data of the entire vehicle is collected through the data acquisition module of the system, the collection situation of the vehicle data is dynamically analyzed to obtain the complexity of the real-time dynamic update of the vehicle data, the boundary threshold of the motor control and data change is analyzed in combination with the analysis of the complexity of the data acquisition state, the motor parameters are analyzed through the vehicle data, and then the interference evaluation boundary is obtained through the error analysis of the vehicle data. The interference evaluation boundary is used to represent the correction of the vehicle data acquisition error, and the adjustable range boundary of the direct control parameters of the motor is analyzed in combination with the boundary threshold. By gradually analyzing, the fitting degree of the motor data is increased, making the range of motor control optimization more reasonable. When the vehicle data changes complexly, the complexity, boundary threshold and adjustable boundary that increase in real time with the vehicle data are used to reduce the dimension of the data, and finally the adaptive optimization control of the motor is performed, greatly improving the accuracy of the adaptive control optimization algorithm for analyzing the changes in the vehicle operating environment, and making the motor control system obtain more accurate control optimization.

[0087] The technical solution of the present invention has been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the protection scope of the present invention.

Claims

1. A motor control optimization method for a vehicle, characterized in that: The control parameters of the motor controller are optimized based on the vehicle data collected by the sensor. The specific analysis process is as follows: Step 1: The vehicle data during the driving process includes all the data generated by the vehicle. The real-time data of the motor operation is obtained through the sensor, and the real-time state of the total vehicle data collection is analyzed to obtain the complexity X corresponding to the real-time state change at the i-th moment. i ; Step 2: The complexity of vehicle data collection in adaptive scenarios corresponding to different complex environments is different. The motor parameters in the vehicle data that actually affect the motor operation control include direct control parameters and indirect parameters. All motor parameters in the vehicle data of motor operation at different times are recorded as a multidimensional state vector, and the boundary threshold of motor control is obtained through multidimensional state vector analysis; Step 3: Perform error interference analysis on the vehicle data collected by the sensor during motor operation to obtain the interference assessment boundary of data collection, and then calculate the adjustable range boundary of the dynamic drive of the motor's direct controllable range in combination with different adaptive scenarios; Step 4: Determine the particle search range through the adjustable range boundary of the motor drive and the interference assessment boundary of the data acquisition, optimize the parameters through the particle swarm search algorithm, transform the nonlinear relationship, and determine the optimal parameters of the system control algorithm through the optimal position of the search; Step 5: Use the neural network algorithm to optimize the control parameters of the motor, build a dynamic model of the vehicle system and its components, and combine the vehicle speed, engine speed, battery state of charge, SOC vehicle status signal and neural network to predict the vehicle speed and torque.

2. The motor control optimization method for a vehicle according to claim 1, characterized in that: The vehicle data of the motor is the data collected for the operation of the whole vehicle, which is recorded as historical operation data and real-time data from the time dimension. The detection indexes corresponding to all change parameters in the motor operation adaptive scenario are extracted and recorded as y 1i ,y 2i ,y 3i ,…y ni , n i represents the number of detection indexes that change with the adaptive scenario in the vehicle data, and i represents the extraction moment in the time dimension. The real-time state analysis is carried out on the vehicle data collected at different moments. The total number of detection indexes in the vehicle data is N, and the number of changes in the detection indexes corresponding to the adaptive scenario at different moments is recorded as n i , and then the change of the detection indexes corresponding to each detection moment i is analyzed to obtain the index change ratio of the real-time change of data collection Then, according to the index change, the complexity X of the real-time change of the vehicle data corresponding to different motor adaptive scenarios is calculated i , and the ratio of the change of the detection index data is calculated k ∈ [1, n i , n i < N, where and respectively represent the change values of the detection index in the adaptive scenario at the i-th moment and the detection index at the (i - 1)-th moment for the same detection index 3. The motor control optimization method for a vehicle according to claim 1, characterized in that: In the second step, the number of parameters of the motor parameters in different adaptive scenarios changes dynamically, and the corresponding adjustment range changes accordingly. The characteristics of the dynamic change boundary of the data related to the motor are dynamically analyzed according to the historical operation data in the vehicle data to obtain the boundary threshold of the adaptive scenario. The number of indirect parameters related to the motor corresponding to the detection index change at the i-th moment is recorded as G1, and the number of detection index changes of the motor's direct control parameters is recorded as G2. The complexity is X i , a point set is obtained according to the analysis of the boundary control equation, which includes the control data of the motor parameters. Then, the elliptic partial differential equation of the dynamic boundary is established according to the point set, and the extreme value of the elliptic differential equation is analyzed with the point set as the value range. Then, the boundary threshold f=max(F) is calculated according to the extreme value. The equation is as follows: Where F is the elliptic function of the dynamic boundary feature, X i is an auxiliary variable, that is, the complexity of data changes at the i-th moment, t i is the time variation function obtained based on historical data, P is the variation function of the corresponding characteristics of the dynamic boundary, t i The and P are obtained based on the historical data of the motor operation, and then the extreme values ​​are solved for the elliptic partial differential equation and the boundary threshold is obtained by using the extreme values ​​for analysis.

4. The motor control optimization method for a vehicle according to claim 1, characterized in that: In step 3, the interference assessment boundary is obtained by performing error assessment on the vehicle data collected by the sensor. In an adaptive scenario, the sensor collects the operation data of the vehicle. The transmission signals of different sensors affect each other in the same adaptive scenario. The vehicle data collected by all sensors are analyzed in layers. The probability that k sensors are assigned to the same group in the jth layer follows a quadratic distribution. P(N,k) i is the probability that k sensors in the i-th layer are assigned to the same group, N is the number of detection indicators, M is the total number of sensors, p is the probability of being assigned to the i-th layer, (1-p) is the probability of not being assigned to the i-th layer, and then the sensor data is analyzed for error based on the binomial distribution, the confidence interval of the data is calculated, the confidence level of the vehicle data is determined, and finally the interference assessment boundary ε of the error is calculated.

5. The motor control optimization method for a vehicle according to claim 1, characterized in that: As the vehicle operating environment and motor operating parameters change, the most adaptive control parameters of the motor change. The adjustable range boundaries of the dynamic drive corresponding to different adaptive scenarios of the motor operation are calculated. The control range of the direct control parameters of the motor operation control is obtained by training the historical operation data with the neural network algorithm. The input vector of the system is u d (t)=[u d1 ,u d2 ,u d3 ,...,u dp ] T , the output is a q-dimensional vector At the same time d (t) The relationship is The operator that represents the relationship between the input vector and the output vector is used to determine the neural function f(x,θ (h) ) approaches the boundary threshold f, ε is the corresponding point value of the interference assessment boundary, h is the data change dimension in the adaptive scene, x is the state vector of the neural network algorithm analysis process that directly controls the parameters, and the adjustable range boundary is [fy i ,f+y i ], where f is the boundary threshold, y i is the value of the motor direct control parameter, satisfying the following performance index function:

6. The motor control optimization method for a vehicle according to claim 1, characterized in that: In the adjustable range of the motor [fy i ,f+y i ], the direct control parameters of the motor are further optimized by using the control strategy of the particle swarm algorithm in the neural network system, where f is the boundary threshold and y i is the value of the motor direct control parameter, and the particle population is described as x = (x1, x2, x3, ..., x m ) T There are m particles in the multidimensional target search space. The position of the particle is determined in the search space. The individual extreme value and group extreme value of the i-th particle are obtained by calculating the objective function, which are recorded as P i and P g , P i =(P i1 ,P i2 ,P i3 ,...,P im ) T , P g =(P g1 ,P g2 ,P g3 ,...,P gm ) T According to the principle that each particle compares its own fitness value with the individual extreme value and the group extreme value after each update, it is continuously updated to determine the optimal fitness position of the group and the individual.

7. The motor control optimization method for a vehicle according to claim 1, characterized in that: The direct control parameters of the motor include speed and torque. During the optimization process of the motor control parameters, the vehicle data of the entire vehicle is dynamically analyzed, and then the control parameters of the motor are optimized based on the particle swarm algorithm in the neural network algorithm.

8. A motor control optimization system for a vehicle, applied in any one of claims 1-7, characterized in that: The system includes a control module, a data acquisition module, a boundary analysis module, a parameter optimization module, and a dynamic adaptive module. The control module is the core of the algorithm. It realizes intelligent control of the motor by sending different control signals to the vehicle's motor and coordinates the joint control of various modules in the vehicle. The data acquisition module includes a speed sensor, a pressure sensor, a humidity sensor, and a temperature sensor. It collects the current signal of the motor under different adaptive scenarios, and then converts the different signals into the form of vehicle data for storage. The boundary analysis module, the dynamic adaptive module, and the parameter optimization module optimize and analyze the motor operation data.

Images