Estimation method and device for driving performance and power generation capacity of motor and storage medium

Abstract

The embodiment of the invention relates to the technical field of vehicles, in particular to a method and device for estimating the driving performance and the power generation capacity of a motor and a storage medium, and the method comprises the steps: obtaining characteristic variables related to the driving performance and the power generation capacity of a vehicle motor; inputting the characteristic variables into a target neural network model to obtain a first maximum driving torque and a first maximum power generation torque of the vehicle motor; a preset record table is obtained, a second maximum driving torque and a second maximum power generation torque of the vehicle motor are obtained based on the preset record table, and the corresponding relation between the vehicle input parameters and the safety limit values is stored in the preset record table; a target maximum driving torque is determined based on the first maximum driving torque and the second maximum driving torque, and a target maximum power generation torque is determined based on the first maximum power generation torque and the second maximum power generation torque. By applying the technical scheme of the invention, the estimation error of the torque data can be reduced.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, specifically to a method, apparatus, and storage medium for estimating the driving performance and power generation capacity of an electric motor. Background Technology

[0002] Torque safety requires real-time monitoring and estimation of the drive motor's real-time driving and power generation capabilities. Currently, the mainstream approach uses a minimum arbitration mechanism, which calculates by looking up fixed parameters in a table or by using a simplified physical model. However, this approach has drawbacks: the model is simplified / linear, resulting in poor dynamic performance; under certain operating conditions, the estimation accuracy is insufficient, and there is significant lag; and the data is only written at the factory, making it unable to adapt to changes throughout the vehicle's lifecycle, such as aging.

[0003] While the method of estimating the real-time driving and power generation capabilities of a motor using a simple neural network can improve accuracy, it also presents challenges in this engineering application, such as difficulties in functional safety verification and insufficient handling of discrete constraints. Summary of the Invention

[0004] In view of the above problems, embodiments of the present invention provide a method, apparatus and storage medium for estimating the driving performance and power generation capacity of a motor, which solves the problems of functional safety verification difficulties and insufficient discrete constraint processing in the existing methods for estimating the real-time driving and power generation capacity of a motor.

[0005] According to one aspect of the present invention, a method for estimating the driving performance and power generation capacity of a motor is provided. The method includes: acquiring feature variables related to the driving performance and power generation capacity of a vehicle motor; inputting the feature variables into a target neural network model to obtain a first maximum driving torque and a first maximum power generation torque of the vehicle motor; acquiring a preset record table and obtaining a second maximum driving torque and a second maximum power generation torque of the vehicle motor based on the preset record table, wherein the preset record table stores the correspondence between vehicle input parameters and safety limits; determining a target maximum driving torque based on the first maximum driving torque and the second maximum driving torque, and determining a target maximum power generation torque based on the first maximum power generation torque and the second maximum power generation torque, wherein the target maximum driving torque is used to estimate the driving performance of the vehicle motor, and the target maximum power generation torque is used to estimate the power generation capacity of the vehicle motor.

[0006] According to another aspect of the present invention, an apparatus for estimating the driving performance and power generation capacity of a motor is provided. The apparatus includes: a first acquisition module, configured to acquire feature variables related to the driving performance and power generation capacity of a vehicle motor; a first obtaining module, configured to input the feature variables into a target neural network model to obtain a first maximum driving torque and a first maximum power generation torque of the vehicle motor; a second obtaining module, configured to acquire a preset record table and obtain a second maximum driving torque and a second maximum power generation torque of the vehicle motor based on the preset record table, wherein the preset record table stores the correspondence between vehicle input parameters and safety limits; and a determination module, configured to determine a target maximum driving torque based on the first maximum driving torque and the second maximum driving torque, and to determine a target maximum power generation torque based on the first maximum power generation torque and the second maximum power generation torque, wherein the target maximum driving torque is used to estimate the driving performance of the vehicle motor, and the target maximum power generation torque is used to estimate the power generation capacity of the vehicle motor.

[0007] According to another aspect of the present invention, a computer device is provided, including: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus; The memory is used to store at least one executable instruction that causes the processor to perform an operation on the method for estimating the driving performance and power generation capacity of the motor in the first aspect.

[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein at least one executable instruction is stored in the storage medium, the executable instruction causing a computer device / apparatus to perform the operation of the method for estimating the driving performance and power generation capacity of a motor of the first aspect.

[0009] The technical solution provided by this invention has the following advantages: This invention obtains feature variables related to the driving performance and power generation capacity of a vehicle motor, then inputs these feature variables into a target neural network model to obtain the first maximum driving torque and the first maximum power generation torque of the vehicle motor. Based on a preset record table, the second maximum driving torque and the second maximum power generation torque of the vehicle motor are obtained. Then, based on the first and second maximum driving torques, the target maximum driving torque is determined, and based on the first and second maximum power generation torques, the target maximum power generation torque is determined. In this way, the feature variables are processed in two ways (i.e., the neural network model and the traditional query based on the preset record table), and then the results of the two processing methods are combined. This not only utilizes the high precision advantage of the neural network but also ensures basic safety boundaries, guaranteeing the functional safety of the system. As a result, the estimation errors of the final target maximum driving torque and target maximum power generation torque are small, and the data reliability is improved.

[0010] The above description is merely an overview of the technical solutions of the embodiments of the present invention. In order to better understand the technical means of the embodiments of the present invention and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0011] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a first embodiment of a method for estimating the driving performance and power generation capacity of an electric motor provided by the present invention is shown. Figure 2 A flowchart illustrating a second embodiment of a method for estimating the driving performance and power generation capacity of an electric motor provided by the present invention is shown. Figure 3 A flowchart illustrating a third embodiment of a method for estimating the driving performance and power generation capacity of an electric motor provided by the present invention is shown. Figure 4 A schematic diagram of the structure of a device for estimating the driving performance and power generation capacity of an electric motor provided by the present invention is shown. Figure 5 A schematic diagram of an embodiment of a computer device provided by the present invention is shown. Detailed Implementation

[0012] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. Although exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein.

[0013] Figure 1 A flowchart illustrating a first embodiment of a method for estimating the driving performance and power generation capacity of an electric motor according to the present invention is shown, the method being executed by a vehicle. Figure 1 As shown, the method includes the following steps: Step S101: Obtain the feature variables related to the driving performance and power generation capacity of the vehicle motor.

[0014] Specifically, the vehicle will acquire characteristic variables related to the driving performance and power generation capacity of the vehicle motor in real time. In this embodiment of the invention, nine real-time dynamic parameters will be acquired, all collected through the vehicle's CAN network or sensors, with a step size of 10ms, covering the three core modules of the motor, motor controller, and battery, specifically including: Motor characteristics: motor speed (n_motor), motor torque command (Tq_cmd), motor winding temperature (Temp_motor).

[0015] Motor controller features: DC bus voltage (V_dc), DC bus current (I_dc), IGBTNTC temperature (Temp_inv).

[0016] Battery characteristics (from BMS): Battery state of charge (Batt_SOC), maximum allowable charging power (P_batt_max_chg), maximum allowable discharging power (P_batt_max_dis).

[0017] Step S102: Input the feature variables into the target neural network model to obtain the first maximum driving torque and the first maximum generating torque of the vehicle motor.

[0018] Specifically, a three-layer hybrid neural network model can be built based on the Python + Matlab deep learning toolbox. Its core is used to learn the complex nonlinear relationships in electric drive systems, achieving high-precision torque estimation, as detailed below: Input layer: 9 feature variables. The minimum and maximum physical boundary values ​​of each feature variable are obtained. Then, each feature variable undergoes min-max normalization: Xnorm = (feature variable...) / (...) Xmin) / (Xmax) Xmin), Xmax / Xmin are fixed to the minimum and maximum physical boundary values ​​of the controller.

[0019] Intermediate hidden layers: Two-layer structure (LSTM layer with 32 neurons + fully connected layer with 16 neurons). The LSTM layer is responsible for learning the temporal dependencies between features, and the fully connected layer implements non-linear combination transformations. Intermediate layer bias initialization: b1 and b2 are both initialized to 0.

[0020] Activation functions: LSTM layers use tanh+sigmoid (standard architecture matching), and fully connected layers use ReLU because it is computationally efficient and can alleviate gradient vanishing. While introducing nonlinearity, it ensures the efficiency and stability of model training, and its overall performance is better than other activation functions.

[0021] Output layer: 2 neurons, which output the first maximum driving torque (Tmax_d) and the first maximum generating torque (Tmax_g) within the next x milliseconds.

[0022] Loss function: MSELoss (This loss function emphasizes severe penalties for large deviations in torque estimation, conforms to the design guidelines for functional safety of electric drive systems, and also considers noise impact and robustness requirements).

[0023] It should be noted that when selecting a neural network model, the number and types of relevant input features (neurons) and output layer nodes (neurons) are determined based on the estimation target of driving power generation (referring to the torque value of driving and power generation capacity, which is the final output result of the neural network). The number of intermediate hidden layers and neurons in the neural network are designed according to the driving power generation estimation accuracy required by technical conditions or project development requirements. Based on the technical conditions or project development requirements, the required estimation accuracy (e.g., ±3%) can be obtained. Here, it refers to the accuracy requirement for torque estimation in the technical development requirements, which is the percentage error between the result calculated by the neural network model and the actual output value. Then, the results are optimized through actual model experiments to control model complexity, avoid overfitting, and meet the accuracy requirements.

[0024] In addition, the neural network model selected after training is referred to here as the target neural network model.

[0025] Step S103: Obtain a preset record table, and obtain the second maximum driving torque and the second maximum generating torque of the vehicle motor based on the preset record table. The preset record table stores the correspondence between vehicle input parameters and safety limits.

[0026] Specifically, the first step is to obtain a preset record table that stores the correspondence between vehicle input parameters and safety limits. This table is generated through extensive bench testing and calibration and is embedded in the controller chip. It mainly includes: a MAP table of external characteristics of motor thermal protection, a MAP table of external characteristics of electronic control thermal protection, a MAP table of battery power limit (associated with SOC, voltage and torque limit corresponding to maximum charge and discharge power), a MAP table of external characteristics of normal output of electric drive system, and a MAP table of torque limit of energy recovery mode.

[0027] These preset record tables are mainly used to store the correspondence between vehicle input parameters and safety limits. For example, the MAP table for electronic control derating protection records the safety output boundaries of core components such as IGBTs.

[0028] The traditional multi-constraint rule arbitration is adopted. By querying the preset record table, each safety boundary limit is obtained, and the minimum value is taken as the conservative safety limit to obtain the second maximum driving torque and the second maximum generating torque of the vehicle motor.

[0029] Step S104: Determine the target maximum driving torque based on the first maximum driving torque and the second maximum driving torque, and determine the target maximum generating torque based on the first maximum generating torque and the second maximum generating torque, wherein the target maximum driving torque is used to estimate the driving performance of the vehicle motor, and the target maximum generating torque is used to estimate the generating capacity of the vehicle motor.

[0030] Specifically, by "integrating high precision of neural networks with high safety of rule arbitration", the first maximum driving torque and the second maximum driving torque, as well as the first maximum generating torque and the second maximum generating torque, are compared. The target maximum driving torque T_Final_d and the target maximum generating torque T_Final_g are obtained by using the rule arbitration result as the absolute safety baseline and the neural network result as only an "optimization supplement".

[0031] For example, in step 1: set the "safety threshold range": based on the second maximum driving torque (Tmax_dt), set a safe range that allows the neural network results to fluctuate, for example [Tmax_dt×0.9,Tmax_dt]; Step 2: Determine whether the first torque value (Tmax_d) falls within the safe range: If Tmax_d∈[Tmax_dt×0.9,Tmax_dt]: the target maximum driving torque T_Final_d=Tmax_d; If Tmax_d > Tmax_dt: directly set T_Final_d = Tmax_dt; If Tmax_d < Tmax_dt × 0.9: take T_Final_d = (Tmax_d + Tmax_dt) / 2.

[0032] The calculation method for the target maximum power generation torque T_Final_g is the same and will not be repeated here.

[0033] Of course, a three-level mechanism of "confidence assessment + system risk assessment + dynamic weight arbitration" can also be used to process the first maximum driving torque and the second maximum driving torque, as well as the first maximum generating torque and the second maximum generating torque, to obtain the target maximum driving torque T_Final_d and the target maximum generating torque T_Final_g, which will be explained in subsequent embodiments.

[0034] It is understood that the target maximum driving torque is used in this embodiment of the invention to estimate the driving performance of the vehicle motor, and the target maximum generating torque is used in this embodiment of the invention to estimate the generating capacity of the vehicle motor.

[0035] This invention obtains feature variables related to the driving performance and power generation capacity of a vehicle motor, then inputs these feature variables into a target neural network model to obtain the first maximum driving torque and the first maximum power generation torque of the vehicle motor. Based on a preset record table, the second maximum driving torque and the second maximum power generation torque of the vehicle motor are obtained. Then, based on the first and second maximum driving torques, the target maximum driving torque is determined, and based on the first and second maximum power generation torques, the target maximum power generation torque is determined. In this way, the feature variables are processed in two ways (i.e., the neural network model and the traditional query based on the preset record table), and then the results of the two processing methods are combined. This not only utilizes the high precision advantage of the neural network but also ensures basic safety boundaries, guaranteeing the functional safety of the system. As a result, the estimation errors of the final target maximum driving torque and target maximum power generation torque are small, and the data reliability is improved.

[0036] As an optional embodiment, before step S102, the method further includes: Step a1: Obtain a sample of characteristic variables related to the driving performance and power generation capacity of the vehicle motor, as well as the third maximum driving torque and the third maximum power generation torque obtained based on the sample of characteristic variables. Step a2: Based on the physical characteristics of the feature variable samples, construct a first nonlinear functional relationship between the maximum driving torque and the feature variable samples; based on the physical characteristics of the feature variable samples, construct a second nonlinear functional relationship between the feature variable samples and the maximum generating torque. Step a3: Based on the first nonlinear function relationship, the second nonlinear function relationship, the third maximum driving torque, and the third maximum generating torque, the parameters of the initial neural network model are adjusted to obtain the target neural network model.

[0037] Specifically, through bench testing, test points were scanned and interpolated across the entire speed range (at speed intervals of 250 rpm), the entire torque range (at torque intervals of 5 Nm), typical voltage points (minimum / maximum / rated), and typical battery SOC (15%, 50%, 90%) to obtain nine feature variable samples related to the driving performance and power generation capacity of the vehicle motor under various operating conditions, as well as the output third maximum driving torque Tmax_d_r and third maximum power generation torque Tmax_g_r. In other words, the feature variable samples and the corresponding third maximum driving torque Tmax_d_r and third maximum power generation torque Tmax_g_r are used as training sample sets and validation sample sets.

[0038] Here, the experimental data recorded above can be divided into 85% as the model training group and 15% as the effect verification group; through training and verification, a suitable neural network estimation model can be finally obtained.

[0039] The relationship between the input and output layer neurons is constructed. Here, the power flow of each input feature variable is derived using a traditional discrete mathematical formula model. The battery output power P_batt_max_dis = U_dc_max * I_dc_max, simplifying the highest voltage model to U_dc_max ≈ k * Batt_SOC + Uo (actually a positively correlated nonlinear relationship), where Uo is the reference voltage, i.e., the theoretical voltage value when SOC = 0, and k is the proportional coefficient. The maximum mechanical power P_mech_max = Tmax_d * (2π * n_motor / 60); the energy conversion efficiency η = P_mech_max / P_batt_max_dis and η can be obtained through bench calibration; the power loss equation is P_loss = P_loss_motor + P_loss_inv; the nonlinear motor loss P_loss_motor and control loss P_loss_inv are simplified to a temperature rise model ΔTemp_motor ≈ P_loss_motor * R_motor, ΔTemp_inv ≈ P_loss_inv * R_inv, where R is the thermal resistance, which can be obtained through bench calibration; thus, P_batt_max_dis = P_mech_max + P_loss_max. The derivation under the power generation mode is similar and will not be elaborated further.

[0040] Due to the complex linear and nonlinear functional relationships described above, the maximum driving torque Tmax_d of the electric drive is affected by the battery discharge capacity and the electric drive output capacity. The following nonlinear functional relationships are derived: Battery discharge capacity = f(Batt_SOC, P_batt_max_dis, V_dc), Electric drive output capacity = f(n_motor, Tq_cmd, I_dc, Temp_motor, Temp_inv); therefore, Tmax_d and the input feature variables can form the following nonlinear functional relationship: Tmax_d = f{f(battery discharge capacity), f(electric drive output capacity)}. Similarly, the maximum generated torque Tmax_g is affected by the battery charging capacity and the electric drive generation status, and the following nonlinear function relationship is constructed: battery charging capacity = f(Batt_SOC, P_batt_max_chg, V_dc), electric drive generation capacity f(n_motor, Tq_cmd, I_dc, Temp_motor, Temp_inv), therefore Tmax_g = f{f(battery charging capacity), f(electric drive generation capacity)}.

[0041] The complexity and mapping relationships of the aforementioned functions are key to neural network learning. Based on the above formulas, we can deduce Tmax_d = P_mech_max / (2π * n_motor / 60). Since P_mech_max is related to other parameters in the system, we group the relationships with these functions: one group represents the battery discharge capability f(Batt_SOC, P_batt_max_dis, V_dc), and the other group represents the electric drive output capability = f(n_motor, Tq_cmd, I_dc, Temp_motor, Temp_inv). This establishes a nested association between Tmax_d = f{f (battery discharge capability), f (electric drive output capability)} and the input feature variables. Similarly, the derivation in the power generation mode yields the relationship between Tmax_g and power generation-related functions. With this input-output correlation analysis, the initial neural network model learns the aforementioned nonlinear function relationships and nested associations during model training, resulting in a learned intermediate neural network model.

[0042] The feature variable samples of the training group are then input into the intermediate neural network model. The initial neuron weights w and neuron biases b1 and b2 are weighted and summed. The neural network prediction results are then obtained through ReLU nonlinear calculation, resulting in the fourth maximum driving torque and the fourth maximum generating torque.

[0043] The error between the prediction results of the neural network and the output values ​​of the corresponding training group is calculated using the loss function MSELoss (i.e., the first error between the fourth maximum driving torque and the third maximum driving torque, and the second error between the fourth maximum generating torque and the third maximum generating torque). The error is calculated through backpropagation, and the weight w and the bias values ​​b1 and b2 of each neuron are adjusted. This training process is continuously repeated to strengthen the training process until the first error and the second error are both less than the preset threshold (e.g., 10%), and the target neural network model is obtained after training.

[0044] In this embodiment of the invention, the neural network is responsible for learning the complex nonlinear characteristics of the system, ensuring that the computational accuracy of the trained target neural network model is higher.

[0045] Figure 2 A flowchart illustrating a second embodiment of a method for estimating the driving performance and power generation capacity of an electric motor according to the present invention is shown, the method being executed by a vehicle. Figure 2 As shown, the method includes the following steps: Step S201: Obtain the feature variables related to the driving performance and power generation capacity of the vehicle's motor. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0046] Step S202: Input the feature variables into the target neural network model to obtain the first maximum driving torque and the first maximum generating torque of the vehicle motor. For details, please refer to [link to details]. Figure 1 Step S102 of the illustrated embodiment will not be described again here.

[0047] Step S203: Based on the feature variables and the preset record table, the second maximum driving torque and the second maximum generating torque of the vehicle motor are obtained. The preset record table stores the correspondence between vehicle input parameters and safety limits.

[0048] Specifically, step S203 includes: Step S2031: Based on the preset record table, obtain the maximum torque under motor thermal protection, the maximum torque under electronic control thermal protection, the maximum output capacity of electric drive, and the maximum torque output by the motor external characteristic curve. Step S2032: Based on the maximum torque under motor thermal protection, the maximum torque under electronic control thermal protection, the maximum output capacity of electric drive, and the maximum torque output from the motor external characteristic curve, the second maximum drive torque is obtained. Step S2033: Based on the preset record table, obtain the maximum power generation torque limit corresponding to the maximum allowable charging power of the battery, the maximum power generation torque limit of the motor under thermal protection constraints, the maximum power generation torque limit of the electronic control under thermal protection constraints, and the maximum allowable power generation torque limit of the vehicle under different energy recovery modes. Step S2034: Based on the maximum generating torque limit corresponding to the maximum allowable charging power of the battery, the maximum generating torque limit of the motor under thermal protection constraints, the maximum generating torque limit of the electronic control under thermal protection constraints, and the maximum allowable generating torque limit of the vehicle under different energy recovery modes, the second maximum generating torque is obtained.

[0049] Specifically, four types of limits are obtained by checking the preset record table: the maximum torque of the motor thermal protection at the current speed (T_motor), the maximum torque of the electronic control thermal protection (T_inv), the maximum torque corresponding to the battery limit (T_bat), and the maximum torque of the normal output of the motor external characteristic curve (T_map).

[0050] The second maximum driving torque Tmax_dt is obtained by comparing T_motor, T_inv, T_bat, and T_map, for example, Tmax_dt = (T_motor + T_inv + T_bat + T_map) / 4. Preferably, Tmax_dt = min(T_motor, T_inv, T_bat, T_map).

[0051] The preset record table is consulted to obtain four types of limits: the torque corresponding to the battery's allowable charging under the current operating conditions (T_bat_limit), the maximum generating torque limit of the motor under thermal protection constraints (T_motor_limit), the maximum generating torque limit of the electronic control under thermal protection constraints (T_inv_limit), and the maximum allowable generating torque limit of the vehicle under different energy recovery modes (T_model_limit).

[0052] The second maximum generating torque, Tmax_gt, is obtained by comparing T_motor_limit, T_inv_limit, T_bat_limit, and T_model_limit. For example, Tmax_gt = min(T_motor_limit + T_inv_limit + T_bat_limit + T_model_limit) / 4. Preferably, Tmax_gt = min(T_motor_limit, T_inv_limit, T_bat_limit, T_model_limit).

[0053] Step S204: Based on the first maximum driving torque and the second maximum driving torque, determine the target maximum driving torque; based on the first maximum generating torque and the second maximum generating torque, determine the target maximum generating torque. The target maximum driving torque is used to estimate the driving performance of the vehicle motor, and the target maximum generating torque is used to estimate the generating capacity of the vehicle motor. For details, please refer to [link to details]. Figure 1 Step S104 of the illustrated embodiment will not be described again here.

[0054] In this embodiment of the invention, a conservative capability limit is output based on the system security boundary to ensure the basic security boundary and guarantee the functional safety of the system.

[0055] Figure 3 A flowchart illustrating a third embodiment of a method for estimating the driving performance and power generation capacity of an electric motor according to the present invention is shown, the method being executed by a vehicle. Figure 3 As shown, the method includes the following steps: Step S301: Obtain the feature variables related to the driving performance and power generation capacity of the vehicle's motor. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0056] Step S302: Input the feature variables into the target neural network model to obtain the first maximum driving torque and the first maximum generating torque of the vehicle motor. For details, please refer to [link to relevant documentation]. Figure 1 Step S102 of the illustrated embodiment will not be described again here.

[0057] Step S303: Based on the feature variables and a preset record table, the second maximum driving torque and the second maximum generating torque of the vehicle motor are obtained. The preset record table stores the correspondence between vehicle input parameters and safety limits. For details, please refer to [link to relevant documentation]. Figure 1 Step S103 of the illustrated embodiment will not be described again here.

[0058] Step S304: Determine the target maximum driving torque based on the first maximum driving torque and the second maximum driving torque; determine the target maximum power generation torque based on the first maximum power generation torque and the second maximum power generation torque.

[0059] Specifically, step S304 includes: Step S3041: Obtain the feature anomaly degree between the feature variable and the feature variable sample; Step S3042: Obtain the first error and the second error obtained within multiple preset periods; Step S3043: Based on the feature anomaly degree, the feature anomaly degree threshold, the first error, the second error, and the error threshold, the confidence weight is obtained; Step S3044: Obtain the maximum value of the motor temperature and the inverter temperature, the temperature safety threshold of the motor temperature and the inverter temperature, and the temperature limit threshold of the motor temperature and the inverter temperature. Step S3045: Based on the maximum value of the motor temperature and the inverter temperature, the temperature safety threshold, and the temperature limit threshold, obtain the temperature risk coefficient of the motor temperature and the inverter temperature. Step S3046: Determine the target maximum driving torque based on the confidence weight, temperature risk coefficient, first maximum driving torque, and second maximum driving torque; determine the target maximum power generation torque based on the confidence weight, temperature risk coefficient, first maximum power generation torque, and second maximum power generation torque.

[0060] Specifically, the output values ​​Tmax_d and Tmax_g of the neural network are subjected to security arbitration with the rule-based arbitration output values ​​Tmax_dt and Tmax_gt. Steps such as calculation result confidence, system risk assessment, and dynamic weight arbitration are introduced to further ensure that the entire system can both output efficiently and have security redundancy. Specifically: Confidence assessment: Calculating confidence weights based on a neural network model. (1) Obtain the feature anomaly degree between feature variables and feature variable samples: The training dataset records the reasonable combination of 9 feature variables under normal vehicle operating conditions, forming a "typical feature distribution". Then, the deviation between the vector composed of the 9 feature variables currently input to the neural network and the "typical feature distribution" used during model training is quantified to obtain the feature anomaly degree.

[0061] (2) The 95th percentile of Mahalanobis distance is used, that is, the Mahalanobis distances of all feature vectors in the training data are sorted, and the maximum Mahalanobis distance of the top 95% of the data is taken as the feature anomaly threshold.

[0062] (3) Obtain the first error = |fourth maximum driving torque - third maximum driving torque| and the second error = |fourth maximum generating torque - third maximum generating torque| within multiple preset cycles (e.g., 50-100 cycles). Obtain the sum of the first error and the second error as the recent error value.

[0063] (4) Obtain the error threshold. The error threshold is set according to the system accuracy requirements, such as the estimation accuracy required mentioned in the previous embodiment, such as ±3%.

[0064] Then, based on the feature anomaly degree, the feature anomaly degree threshold, the first error, the second error, and the error threshold, the confidence weight is obtained: C_n = w1 * C_input + w2 * C_history, where the weight coefficients w1 + w2 = 1, and C_n ranges from [0, 1]. Input confidence C_input = 1 - min(1, |feature anomaly| / feature anomaly threshold); historical confidence C_history = 1 - min(1, recent error value / error threshold), where C_n = 1 (completely reliable) and C_n = 0 (completely unreliable). w1 corresponds to the input feature confidence weight, and w2 corresponds to the historical prediction confidence weight. It should be noted that if the recent error value differs significantly from the error threshold, w2 should be set smaller; conversely, if the feature anomaly value or the feature anomaly threshold is large, w1 should be set smaller.

[0065] System risk assessment: Screening key influencing factors, namely inverter and motor temperature risks: Temperature risk coefficient = max{0, (current temperature - temperature safety threshold) / (temperature limit threshold - temperature safety threshold)}, where: current temperature: the maximum value between motor temperature and inverter temperature; temperature safety threshold for motor and inverter temperatures: the upper limit of the normal operating temperature of the system design (e.g., 80°C); temperature limit threshold for motor and inverter temperatures: the highest allowable operating temperature of the system (e.g., 120°C). The temperature risk range is [0,1], where 1 represents extremely high risk and 0 represents no risk.

[0066] Then, through dynamic weight arbitration, the confidence weight, temperature risk coefficient, and the first and second maximum driving torques are calculated to obtain the target maximum driving torque; similarly, the confidence weight, temperature risk coefficient, and the first and second maximum power generation torques are calculated to obtain the target maximum power generation torque. This allows the weight of the rule arbitration result to be automatically increased when there are abnormal inputs or insufficient recent model accuracy, ensuring the safety of the estimation.

[0067] Further, step S3046 includes: Step b1: Based on the first maximum driving torque and the second maximum driving torque, obtain the first intermediate maximum driving torque; Step b2: Based on the first maximum driving torque, the second maximum driving torque, and the confidence weight, the second intermediate maximum driving torque is obtained; Step b3: Based on the temperature risk coefficient, the first intermediate maximum driving torque, and the second intermediate maximum driving torque, the target maximum driving torque is obtained. Step b4: Based on the first maximum power generation torque and the second maximum power generation torque, obtain the first intermediate maximum power generation torque; Step b5: Based on the first maximum power generation torque, the second maximum power generation torque, and the confidence weight, the second intermediate maximum power generation torque is obtained; Step b6: Based on the temperature risk coefficient, the first intermediate maximum power generation torque, and the second intermediate maximum power generation torque, the target maximum power generation torque is obtained.

[0068] Optionally, dynamic weighted arbitration is used, with confidence weight β=C_n, α=temperature risk coefficient, calculating the conservative driving value T_base_d=min(Tmax_d,Tmax_dt), calculating the weighted mixed value T_mix_d=β*Tmax_d+(1-β)Tmax_dt, and finally calculating the final arbitration value, the target maximum driving torque T_Final_d=α*T_base_d+(1-α)*T_mix_d.

[0069] Set the confidence weight β=C_n, α=temperature risk coefficient, calculate the conservative power generation value T_base_g=min(Tmax_g,Tmax_gt), calculate the weighted mixed value T_mix_g=β*Tmax_g+(1-β)Tmax_gt, and finally calculate the final arbitration value, the target maximum power generation torque T_Final_g=α*T_base_g+(1-α)*T_mix_g.

[0070] In this embodiment of the invention, by combining confidence weight and temperature risk coefficient, the actual needs of torque estimation for electric drive systems are ensured, and the torque estimation error can be reduced from more than 15% in traditional methods to less than 5%. At the same time, the reliability of the system is ensured through multi-level safety mechanisms.

[0071] Figure 4 A schematic diagram of an embodiment of a device for estimating the driving performance and power generation capacity of a motor according to the present invention is shown. The device includes: The first acquisition module 401 is used to acquire feature variables related to the driving performance and power generation capacity of the vehicle motor; The first module 402 is used to input the feature variables into the target neural network model to obtain the first maximum driving torque and the first maximum generating torque of the vehicle motor. The second obtaining module 403 is used to obtain a preset record table and obtain the second maximum driving torque and the second maximum generating torque of the vehicle motor based on the preset record table. The preset record table stores the correspondence between vehicle input parameters and safety limits. The determining module 404 is used to determine a target maximum driving torque based on a first maximum driving torque and a second maximum driving torque, and to determine a target maximum generating torque based on a first maximum generating torque and a second maximum generating torque, wherein the target maximum driving torque is used to estimate the driving performance of the vehicle motor, and the target maximum generating torque is used to estimate the generating capacity of the vehicle motor.

[0072] This invention obtains feature variables related to the driving performance and power generation capacity of a vehicle motor, then inputs these feature variables into a target neural network model to obtain the first maximum driving torque and the first maximum power generation torque of the vehicle motor. Based on a preset record table, the second maximum driving torque and the second maximum power generation torque of the vehicle motor are obtained. Then, based on the first and second maximum driving torques, the target maximum driving torque is determined, and based on the first and second maximum power generation torques, the target maximum power generation torque is determined. In this way, the feature variables are processed in two ways (i.e., the neural network model and the traditional query based on the preset record table), and then the results of the two processing methods are combined. This not only utilizes the high precision advantage of the neural network but also ensures basic safety boundaries, guaranteeing the functional safety of the system. As a result, the estimation errors of the final target maximum driving torque and target maximum power generation torque are small, and the data reliability is improved.

[0073] In one alternative embodiment, the device further includes: The second acquisition module is used to acquire, before inputting the feature variables into the target neural network model, a sample of feature variables related to the driving performance and power generation capacity of the vehicle motor, as well as a third maximum driving torque and a third maximum power generation torque obtained based on the sample of feature variables. The module is used to construct a first nonlinear functional relationship between the physical characteristics of the feature variable samples and the maximum driving torque, and a second nonlinear functional relationship between the physical characteristics of the feature variable samples and the maximum generating torque. The adjustment module is used to adjust the parameters of the initial neural network model based on the first nonlinear function relationship, the second nonlinear function relationship, the third maximum driving torque, and the third maximum generating torque to obtain the target neural network model.

[0074] In one alternative approach, the module is adjusted, including: The first acquisition unit is used to acquire an initial neural network model and control the initial neural network model to learn the first nonlinear function relationship and the second nonlinear function relationship to obtain an intermediate neural network model; The input unit is used to input the feature variable samples into the intermediate neural network model to obtain the fourth maximum driving torque and the fourth maximum generating torque. The second acquisition unit is used to acquire the first error between the fourth maximum driving torque and the third maximum driving torque, and to acquire the second error between the fourth maximum generating torque and the third maximum generating torque. The adjustment unit is used to adjust the neuron weights and neuron biases in the intermediate neural network model until both the first error and the second error are less than a preset threshold, thus obtaining the target neural network model.

[0075] In one alternative embodiment, the second receiving module 403 includes: The third acquisition unit is used to acquire, based on a preset record table, the maximum torque under motor thermal protection, the maximum torque under electronic control thermal protection, the maximum output capacity of electric drive, and the maximum torque output by the motor external characteristic curve. The first obtaining unit is used to obtain the second maximum driving torque based on the maximum torque under motor thermal protection, the maximum torque under electronic control thermal protection, the maximum output capability of electric drive, and the maximum torque output from the motor external characteristic curve. The fourth acquisition unit is used to acquire, based on a preset record table, the maximum power generation torque limit corresponding to the maximum allowable charging power of the battery, the maximum power generation torque limit of the motor under thermal protection constraints, the maximum power generation torque limit of the electronic control under thermal protection constraints, and the maximum allowable power generation torque limit of the vehicle under different energy recovery modes. The second obtaining unit is used to obtain the second maximum generating torque based on the maximum generating torque limit corresponding to the maximum allowable charging power of the battery, the maximum generating torque limit of the motor under thermal protection constraints, the maximum generating torque limit of the electronic control under thermal protection constraints, and the maximum allowable generating torque limit of the vehicle under different energy recovery modes.

[0076] In one alternative approach, the first obtained unit includes: The first setting subunit is used to obtain the minimum value among the maximum torque under motor thermal protection, the maximum torque under electronic control thermal protection, the maximum output capability of electric drive, and the maximum torque output from the motor external characteristic curve, as the second maximum driving torque.

[0077] In one alternative approach, the second obtaining unit includes: The second setting subunit is used to obtain the minimum value among the maximum power generation torque limit corresponding to the maximum allowable charging power of the battery, the maximum power generation torque limit of the motor under thermal protection constraints, the maximum power generation torque limit of the electronic control under thermal protection constraints, and the maximum allowable power generation torque limit of the vehicle under different energy recovery modes, as the second maximum power generation torque.

[0078] In one alternative approach, determining module 404 includes: The fifth acquisition unit is used to acquire the feature anomaly degree between the feature variable and the feature variable sample; The sixth acquisition unit is used to acquire the first error and the second error obtained within multiple preset periods; The third obtaining unit is used to obtain the confidence weight based on the feature anomaly degree, the feature anomaly degree threshold, the first error, the second error, and the error threshold. The seventh acquisition unit is used to acquire the maximum value of the motor temperature and the inverter temperature, the temperature safety threshold of the motor temperature and the inverter temperature, and the temperature limit threshold of the motor temperature and the inverter temperature. The fourth unit is used to obtain the temperature risk coefficient of motor temperature and inverter temperature based on the maximum value of motor temperature and inverter temperature, temperature safety threshold and temperature limit threshold. The determining unit is used to determine the target maximum driving torque based on the confidence weight, temperature risk coefficient, first maximum driving torque, and second maximum driving torque, and to determine the target maximum generating torque based on the confidence weight, temperature risk coefficient, first maximum generating torque, and second maximum generating torque.

[0079] In one alternative approach, determining the unit includes: The first obtaining subunit is used to obtain the first intermediate maximum driving torque based on the first maximum driving torque and the second maximum driving torque; The second sub-unit is used to obtain the second intermediate maximum driving torque based on the first maximum driving torque, the second maximum driving torque, and the confidence weight. The third subunit is used to obtain the target maximum driving torque based on the temperature risk coefficient, the first intermediate maximum driving torque, and the second intermediate maximum driving torque. The fourth sub-unit is used to obtain the first intermediate maximum power generation torque based on the first maximum power generation torque and the second maximum power generation torque; The fifth sub-unit is used to obtain the second intermediate maximum power generation torque based on the first maximum power generation torque, the second maximum power generation torque, and the confidence weight. The sixth sub-unit is used to obtain the target maximum power generation torque based on the temperature risk coefficient, the first intermediate maximum power generation torque, and the second intermediate maximum power generation torque.

[0080] Figure 5 The diagram illustrates a structural schematic of an embodiment of the computer device of the present invention. The specific embodiments of the present invention do not limit the specific implementation of the computer device.

[0081] like Figure 5 As shown, the computer device may include: a processor 502, a communications interface 504, a memory 506, and a communications bus 508.

[0082] The processor 502, communication interface 504, and memory 506 communicate with each other via communication bus 508. Communication interface 504 is used to communicate with other network elements, such as clients or other servers. The processor 502 executes program 510, specifically performing the relevant steps in the above-described embodiment of the method for estimating the drive performance and power generation capacity of the motor.

[0083] Specifically, program 510 may include program code, which includes computer-executable instructions.

[0084] Processor 502 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The computer device includes one or more processors, which may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0085] Memory 506 is used to store program 510. Memory 506 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0086] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0087] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. Similarly, for the sake of brevity and to aid in understanding one or more aspects of the invention, in the description of exemplary embodiments of the invention above, various features of the embodiments are sometimes grouped together in a single embodiment, figure, or description thereof. The claims, which follow the detailed description, are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of the invention.

[0088] Those skilled in the art will understand that the modules in the device of the embodiment can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiment can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components, except that at least some of such features and / or processes or units are mutually exclusive.

[0089] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be construed as limiting the order of execution.

Claims

1. A method for estimating the driving performance and power generation capacity of an electric motor, characterized in that, The method includes: Obtain the characteristic variables related to the driving performance and power generation capacity of the vehicle's motor; The feature variables are input into the target neural network model to obtain the first maximum driving torque and the first maximum generating torque of the vehicle motor. Obtain a preset record table, and obtain the second maximum driving torque and the second maximum generating torque of the vehicle motor based on the preset record table, wherein the preset record table stores the correspondence between vehicle input parameters and safety limits; Based on the first maximum driving torque and the second maximum driving torque, a target maximum driving torque is determined. Based on the first maximum generating torque and the second maximum generating torque, a target maximum generating torque is also determined. The target maximum driving torque is used to estimate the driving performance of the vehicle motor, and the target maximum generating torque is used to estimate the generating capacity of the vehicle motor.

2. The method according to claim 1, characterized in that, Before inputting the feature variables into the target neural network model, the method further includes: Acquire a sample of characteristic variables related to the driving performance and power generation capacity of the vehicle motor, and obtain the third maximum driving torque and the third maximum power generation torque based on the sample of characteristic variables; Based on the physical characteristics of the feature variable samples, a first nonlinear functional relationship between them and the maximum driving torque is constructed; based on the physical characteristics of the feature variable samples, a second nonlinear functional relationship between them and the maximum generating torque is constructed. Based on the first nonlinear function relationship, the second nonlinear function relationship, the third maximum driving torque, and the third maximum generating torque, the parameters of the initial neural network model are adjusted to obtain the target neural network model.

3. The method according to claim 2, characterized in that, Based on the first nonlinear function relationship, the second nonlinear function relationship, the third maximum driving torque, and the third maximum generating torque, the parameters of the initial neural network model are adjusted to obtain the target neural network model, including: Obtain the initial neural network model, and control the initial neural network model to learn the first nonlinear function relationship and the second nonlinear function relationship to obtain an intermediate neural network model; The feature variable samples are input into the intermediate neural network model to obtain the fourth maximum driving torque and the fourth maximum generating torque; Obtain the first error between the fourth maximum driving torque and the third maximum driving torque, and obtain the second error between the fourth maximum generating torque and the third maximum generating torque; Adjust the neuron weights and biases within the intermediate neural network model until both the first error and the second error are less than a preset threshold to obtain the target neural network model.

4. The method according to claim 1, characterized in that, The step of obtaining a preset record table and obtaining the second maximum driving torque and the second maximum generating torque of the vehicle motor based on the preset record table includes: Based on the preset record table, obtain the maximum torque under motor thermal protection, the maximum torque under electronic control thermal protection, the maximum output capacity of electric drive, and the maximum torque output by the motor external characteristic curve. The second maximum driving torque is obtained based on the maximum torque under the motor thermal protection, the maximum torque under the electronic control thermal protection, the maximum output capacity of the electric drive, and the maximum torque output from the motor external characteristic curve. Based on the preset record table, obtain the maximum power generation torque limit corresponding to the maximum allowable charging power of the battery, the maximum power generation torque limit of the motor under thermal protection constraints, the maximum power generation torque limit of the electronic control under thermal protection constraints, and the maximum allowable power generation torque limit of the vehicle under different energy recovery modes. The second maximum generating torque is obtained based on the maximum generating torque limit corresponding to the maximum allowable charging power of the battery, the maximum generating torque limit of the motor under the thermal protection constraint, the maximum generating torque limit of the electronic control under the thermal protection constraint, and the maximum allowable generating torque limit of the vehicle under different energy recovery modes.

5. The method according to claim 4, characterized in that, The second maximum drive torque is obtained based on the maximum torque under the motor thermal protection, the maximum torque under the electronic control thermal protection, the maximum output capacity of the electric drive, and the maximum torque output from the motor external characteristic curve, including: The minimum value among the maximum torque under motor thermal protection, the maximum torque under electronic control thermal protection, the maximum output capability of the electric drive, and the maximum torque output from the motor external characteristic curve is obtained as the second maximum driving torque.

6. The method according to claim 4, characterized in that, The second maximum generating torque is obtained based on the maximum generating torque limit corresponding to the maximum allowable charging power of the battery, the maximum generating torque limit of the motor under thermal protection constraints, the maximum generating torque limit of the electronic control under thermal protection constraints, and the maximum allowable generating torque limit of the vehicle under different energy recovery modes, including: The minimum value among the maximum allowable charging power of the battery, the maximum generating torque limit of the motor under thermal protection constraints, the maximum generating torque limit of the electronic control under thermal protection constraints, and the maximum allowable generating torque limit of the vehicle under different energy recovery modes is obtained as the second maximum generating torque.

7. The method according to claim 3, characterized in that, The step of determining the target maximum driving torque based on the first maximum driving torque and the second maximum driving torque, and determining the target maximum power generation torque based on the first maximum power generation torque and the second maximum power generation torque, includes: Obtain the feature anomaly degree between the feature variable and the feature variable sample; Obtain the first error and the second error obtained within multiple preset periods; Based on the feature anomaly degree, the feature anomaly threshold, the first error, the second error, and the error threshold, the confidence weight is obtained; Obtain the maximum value of the motor temperature and the inverter temperature, the temperature safety threshold of the motor temperature and the inverter temperature, and the temperature limit threshold of the motor temperature and the inverter temperature. Based on the maximum value of the motor temperature and the inverter temperature, the temperature safety threshold, and the temperature limit threshold, the temperature risk coefficient of the motor temperature and the inverter temperature is obtained. The target maximum driving torque is determined based on the confidence weight, the temperature risk coefficient, the first maximum driving torque, and the second maximum driving torque. The target maximum power generation torque is also determined based on the confidence weight, the temperature risk coefficient, the first maximum power generation torque, and the second maximum power generation torque.

8. The method according to claim 7, characterized in that, The determination of the target maximum driving torque based on the confidence weight, the temperature risk coefficient, the first maximum driving torque, and the second maximum driving torque, and the determination of the target maximum power generation torque based on the confidence weight, the temperature risk coefficient, the first maximum power generation torque, and the second maximum power generation torque, include: Based on the first maximum driving torque and the second maximum driving torque, the first intermediate maximum driving torque is obtained; Based on the first maximum driving torque, the second maximum driving torque, and the confidence weight, the second intermediate maximum driving torque is obtained; The target maximum driving torque is obtained based on the temperature risk coefficient, the first intermediate maximum driving torque, and the second intermediate maximum driving torque. Based on the first maximum power generation torque and the second maximum power generation torque, the first intermediate maximum power generation torque is obtained; Based on the first maximum power generation torque, the second maximum power generation torque, and the confidence weight, the second intermediate maximum power generation torque is obtained; The target maximum power generation torque is obtained based on the temperature risk coefficient, the first intermediate maximum power generation torque, and the second intermediate maximum power generation torque.

9. A device for estimating the driving performance and power generation capacity of an electric motor, characterized in that, The device includes: The first acquisition module is used to acquire feature variables related to the driving performance and power generation capacity of the vehicle motor; The first obtaining module is used to input the feature variables into the target neural network model to obtain the first maximum driving torque and the first maximum generating torque of the vehicle motor. The second obtaining module is used to obtain a preset record table and obtain the second maximum driving torque and the second maximum generating torque of the vehicle motor based on the preset record table, wherein the preset record table stores the correspondence between vehicle input parameters and safety limits; The determining module is configured to determine a target maximum driving torque based on the first maximum driving torque and the second maximum driving torque, and to determine a target maximum generating torque based on the first maximum generating torque and the second maximum generating torque, wherein the target maximum driving torque is used to estimate the driving performance of the vehicle motor, and the target maximum generating torque is used to estimate the generating capacity of the vehicle motor.

10. A computer device, characterized in that, include: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction that causes the processor to perform the operation of the method for estimating the driving performance and power generation capacity of the motor as described in any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, The storage medium stores at least one executable instruction, which, when executed on the storage device / apparatus, causes the computer device / apparatus to perform the operation of the method for estimating the driving performance and power generation capacity of the motor as described in any one of claims 1-8.

Images