Electric vehicle state detection method and device

Abstract

The invention discloses an electric vehicle state detection method and device. The method comprises the steps that current driving feature data of a target vehicle are obtained; respectively inputting the driving characteristic data into a pre-trained data driving model and a pre-trained mechanism model; acquiring a first state detection result output by the data driving model; obtaining a second state detection result output by the mechanism model; determining a first weight value of the first state detection result and a second weight value of the second state detection result according to the driving feature data and the historical driving feature data; and performing weighted summation on the first state detection result and the second state detection result according to the first weight value and the second weight value to obtain a target state detection result. According to the invention, the accuracy and robustness of electric vehicle state detection can be improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent management of electric vehicles, and in particular to a method and device for detecting the status of an electric vehicle. Background Art

[0002] The status detection results of electric vehicles can be well applied to the battery management system and intelligent driving assistance system of electric vehicles, as well as vehicle health monitoring and predictive maintenance, and realize intelligent scheduling and energy consumption management of electric vehicles. The status monitoring results can help drivers understand the vehicle status in real time during driving, optimize driving decisions, ensure driving safety, and improve driving experience.

[0003] Existing electric vehicle status detection solutions rely on mechanism models and data-driven models. Among them, the mechanism model can provide accurate predictions under steady-state conditions based on the physical laws of the vehicle, but it is difficult to adapt to complex working conditions and dynamic changes, and the model accuracy has limitations in practical applications; the data-driven model uses a large amount of sensor data and historical data, and uses deep learning and other algorithms for status prediction. It has strong flexibility, but it is prone to inaccurate or even erroneous output when the data is insufficient or abnormal. The existing electric vehicle status detection solution cannot dynamically balance the advantages and disadvantages of the mechanism model and the data-driven model under different working conditions, resulting in low accuracy and stability in vehicle status detection. Summary of the invention

[0004] The present application provides an electric vehicle state detection method and device, which obtains the current driving characteristic data of the target vehicle; then, the driving characteristic data is respectively input into the pre-trained data-driven model and the mechanism model; then, the first state detection result output by the data-driven model is obtained; and, the second state detection result output by the mechanism model is obtained; secondly, the first weight value of the first state detection result and the second weight value of the second state detection result are determined according to the driving characteristic data and the historical driving characteristic data; finally, the first state detection result and the second state detection result are weighted and summed according to the first weight value and the second weight value to obtain the target state detection result. The weights of the mechanism model and the data-driven model are dynamically adjusted, the flexibility is improved when the data is normal, and the physical rationality is guaranteed when the data is abnormal, thereby improving the accuracy and stability of the state detection.

[0005] In a first aspect, an embodiment of the present application provides a method for detecting the state of an electric vehicle, the method comprising the following steps:

[0006] Acquiring current driving characteristic data of the target vehicle, wherein the driving characteristic data is used to characterize driving parameters of the target vehicle;

[0007] Inputting the driving characteristic data into a pre-trained data-driven model and a mechanism model respectively, wherein the data-driven model is trained based on the historical driving characteristic data of the target vehicle, and the mechanism model is trained based on the physical principle of the target vehicle;

[0008] Obtaining a first state detection result output by the data-driven model; and obtaining a second state detection result output by the mechanism model, wherein the first state detection result includes a plurality of first detection values ​​corresponding to a plurality of state indicators of the target vehicle, and the second state detection result includes a plurality of second detection values ​​corresponding to a plurality of state indicators of the target vehicle;

[0009] Determine a first weight value of the first state detection result and a second weight value of the second state detection result according to the driving characteristic data and the historical driving characteristic data;

[0010] The first state detection result and the second state detection result are weightedly summed according to the first weight value and the second weight value to obtain a target state detection result.

[0011] In a second aspect, an embodiment of the present application provides an electric vehicle state detection device, the device comprising:

[0012] A first acquisition unit, used to acquire current driving characteristic data of the target vehicle;

[0013] An input unit, used to input the driving characteristic data into a pre-trained data-driven model and a mechanism model, respectively, wherein the data-driven model is trained based on the historical driving characteristic data of the target vehicle, and the mechanism model is trained based on the physical principle of the target vehicle;

[0014] A second acquisition unit is used to acquire a first state detection result output by the data-driven model; and acquire a second state detection result output by the mechanism model, wherein the first state detection result includes a plurality of first detection values ​​corresponding to a plurality of state indicators of the target vehicle, and the second state detection result includes a plurality of second detection values ​​corresponding to a plurality of state indicators of the target vehicle;

[0015] a determining unit, configured to determine a first weight value of the first state detection result and a second weight value of the second state detection result according to the driving characteristic data and the historical driving characteristic data;

[0016] A state detection unit is used to perform weighted summation on the first state detection result and the second state detection result according to the first weight value and the second weight value to obtain a target state detection result.

[0017] In a third aspect, an embodiment of the present application provides an electronic device, comprising a processor, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the program includes instructions for executing the steps in the first aspect of the embodiment of the present application.

[0018] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program / instruction stored thereon, and the computer program / instruction is executed by a processor to implement the steps of the method described in the first aspect above.

[0019] It can be seen that in the embodiment of the present application, the current driving characteristic data of the target vehicle is obtained; then, the driving characteristic data is respectively input into the pre-trained data-driven model and the mechanism model; then, the first state detection result output by the data-driven model is obtained; and, the second state detection result output by the mechanism model is obtained; secondly, the first weight value of the first state detection result and the second weight value of the second state detection result are determined according to the driving characteristic data and the historical driving characteristic data; finally, the first state detection result and the second state detection result are weighted and summed according to the first weight value and the second weight value to obtain the target state detection result. Compared with the existing scheme of performing electric vehicle state detection respectively through the mechanism model or the data-driven model, the electric vehicle state detection under different working conditions is realized by dynamically adjusting the weights of the mechanism model and the data-driven model, which is conducive to improving the accuracy and robustness of vehicle state detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 It is a flowchart of the steps of an electric vehicle state detection method provided in an embodiment of the present application;

[0022] Figure 2 This is a system architecture diagram of an electric vehicle status detection system provided in an embodiment of the present application;

[0023] Figure 3 It is an overall flow chart of an electric vehicle state detection method provided in an embodiment of the present application;

[0024] Figure 4 This is a block diagram of the functional units of an electric vehicle state detection device provided in an embodiment of the present application;

[0025] Figure 5 It is a structural block diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0027] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0028] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0029] At present, the existing electric vehicle status detection scheme mainly uses mechanism models and data-driven models to detect vehicle status. The mechanism model has high accuracy but is difficult to adapt to complex working conditions and dynamic changes. The data-driven model has strong flexibility but is prone to inaccurate or even erroneous output when the data is insufficient or abnormal. The existing electric vehicle status detection scheme cannot dynamically balance the advantages and disadvantages of the mechanism model and the data-driven model under different working conditions, and it is difficult to maintain reliable vehicle status detection in complex situations.

[0030] In view of the above problems, an embodiment of the present application provides an electric vehicle status detection method and device, which will be described in detail below in conjunction with the accompanying drawings.

[0031] See also Figure 1 , Figure 1 is a flowchart of a method for detecting the state of an electric vehicle provided in an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:

[0032] Step S110, obtaining current driving characteristic data of the target vehicle, wherein the driving characteristic data is used to characterize driving parameters of the target vehicle.

[0033] In a possible embodiment, the step of obtaining the current driving characteristic data of the target vehicle includes:

[0034] Acquire current driving data of the target vehicle;

[0035] Acquire the multiple status indicators for status detection of the target vehicle, where the status indicators are used to detect the status of the target vehicle from multiple indicator dimensions;

[0036] Determine target driving data corresponding to the multiple state indicators, wherein the target driving data is used to obtain detection results corresponding to the multiple state indicators;

[0037] A target operation is performed on the target driving data to obtain the driving feature data, wherein the target operation includes a preprocessing operation and a feature engineering operation.

[0038] The driving data may come from a vehicle CAN bus, a sensor, an on-board diagnostic system (OBD), a remote monitoring platform, etc. The driving data type may include vehicle information data, battery data, motor data, environmental data, and driving behavior data.

[0039] Among them, the vehicle information data includes vehicle model, battery capacity, motor power, etc.; battery data includes voltage, current, temperature, SOC (state of charge), etc.; motor data includes speed, torque, temperature, etc.; environmental data includes temperature, humidity, road conditions, etc.; driving behavior data includes vehicle speed, acceleration, braking, steering, etc.

[0040] Among them, the status indicators of electric vehicles usually include at least one of multiple indicators such as battery status indicators, power system indicators, driving status indicators, environmental and external conditions indicators, fault detection indicators, safety indicators, user experience indicators and energy consumption indicators. The status indicators can be set exclusively according to user needs and the vehicle's own status, and different vehicles can use different status indicators.

[0041] Exemplarily, the battery status indicator can specifically characterize the battery status of the vehicle from multiple dimensions such as battery power, battery charging status, battery health status, battery temperature, and number of charge and discharge cycles; the power system indicator can specifically characterize the power system operation status of the vehicle from multiple dimensions such as motor power, current and voltage, and motor capacity conversion efficiency; the driving status indicator can specifically characterize the driving status of the vehicle from multiple dimensions such as speed, acceleration, and mileage; the environment and external condition indicators can specifically characterize the environmental status of the vehicle from multiple dimensions such as external temperature and road condition information; the fault detection indicator can specifically characterize the fault status of the vehicle from multiple dimensions such as fault code and abnormal behavior; the safety indicator can specifically characterize the safety status of the vehicle from multiple dimensions such as braking system status and stability control status; the user experience indicator can specifically characterize the driving status of the vehicle from multiple dimensions such as charging time and cruising range; the energy consumption indicator can specifically characterize the energy consumption status of the vehicle from multiple dimensions such as energy consumption rate and regenerative braking energy recovery rate.

[0042] Specifically, the target driving data is preprocessed, including data cleaning, data conversion, data standardization and data normalization; and the target driving data is feature-engineered, including extracting time series features, creating window features and combining features for the data obtained after the preprocessing.

[0043] Among them, the data cleaning includes removing or filling missing values ​​and deleting outliers; data conversion includes converting timestamps into date and time formats and converting categorical variables into numerical variables; data standardization includes scaling numerical data to the same scale, such as using Z-score standardization; data normalization includes scaling data to between 0 and 1, such as using Min-Max standardization.

[0044] Among them, the creation of window features can be speed, acceleration, etc. in the past period of time; feature selection refers to the use of statistical tests, model selection and other methods to select the most useful features for detection; feature combination refers to combining multiple features to create new features.

[0045] Exemplarily, the time series feature extraction of the preprocessed data can be performed for the time series characteristics of sensor data (such as voltage, current, temperature, etc.), and the statistical features within the time window, such as mean, variance, maximum value, minimum value, rate of change, etc., are extracted. By analyzing the trend and fluctuation of the time series data, key features such as battery power changes and energy consumption fluctuations are identified, helping the model capture long-term changes and instantaneous changes.

[0046] It can be seen that in this embodiment, by obtaining multiple state indicators for state detection of the target vehicle, the driving data corresponding to the multiple state indicators are determined, and the driving data corresponding to the multiple state indicators are further preprocessed and feature engineering is performed to obtain the current driving feature data of the target vehicle. Accurately obtaining the driving data required for state detection of the target vehicle is achieved, which is conducive to further processing the driving data according to the detection model to obtain the vehicle state detection result, and realize accurate and robust vehicle state detection.

[0047] Step S120, inputting the driving characteristic data into a pre-trained data-driven model and a mechanism model respectively, wherein the data-driven model is trained based on the historical driving characteristic data of the target vehicle, and the mechanism model is trained based on the physical principle of the target vehicle.

[0048] Among them, the mechanism model is mainly based on the physical laws of the vehicle (such as dynamic model, battery model, driver fatigue model), which can provide accurate predictions under steady-state conditions. However, since the mechanism model is difficult to adapt to complex working conditions and dynamic changes, the model accuracy is limited in practical applications.

[0049] Among them, the mechanism model can include the dynamic model of electric vehicles (describing the physical process of vehicle acceleration, deceleration, and steering), the battery charging and discharging model (reflecting the health status and remaining life of the battery), the temperature mechanism model (such as battery heat conduction and motor heat generation), and the driver's fatigue model (reflecting the driver's fatigue state based on physiological data). These models can ensure that the output has good physical rationality.

[0050] Specifically, the vehicle dynamics model mainly describes the acceleration, deceleration, and turning processes of the vehicle based on Newton's laws of motion, and constructs the dynamics equation through parameters such as vehicle mass, resistance, and acceleration. Taking vehicle acceleration as an example, the dynamics equation is as follows:

[0051] F=m*a=F drive -F drag -F friction

[0052] Among them, F drive is the vehicle driving force, F drag is the air resistance, F friction is the friction force, m is the vehicle mass, and a is the acceleration.

[0053] Specifically, the battery model is mainly based on electrochemical principles, describing the battery's charging and discharging process, health status and attenuation. The main parameters include the battery's internal resistance, capacity and temperature. These parameters are crucial for predicting the remaining life and range of the battery. Taking the battery SOC (remaining charge) as an example, the calculation formula is as follows:

[0054]

[0055] Among them, SOC(t) refers to the remaining battery capacity, SOC(0) is the battery SOC in the initial state, C battery is the rated capacity of the battery, and I(t) is the battery current at time t.

[0056] Specifically, the wear model is mainly based on road friction, tire pressure and vehicle load, and the mechanism model can estimate the wear rate and amount of tire wear. The output is usually the percentage of tire wear or the remaining life. Through braking frequency, braking force and friction coefficient, the mechanism model can predict the wear state of the brake pad, and the output can be the wear amount or remaining service life of the brake pad. Taking the wear amount as an example, the calculation formula is as follows:

[0057] W wear = k * F friction *d

[0058] Among them, W wear is the wear amount, k is the wear coefficient of the material, F friction is the friction force and d is the distance traveled.

[0059] Specifically, the fatigue model mainly combines driving time and vehicle behavior data to build a vehicle fatigue prediction model. This model can help the system determine the fatigue state of the vehicle. Taking fatigue life as an example, the calculation formula is as follows:

[0060]

[0061] Among them, N f is the fatigue life of the component (i.e. the number of fatigue cycles it can withstand), σ max is the maximum stress, σ fatigue is the fatigue limit of the material, and k is the fatigue exponent of the material.

[0062] The data-driven model mainly detects the health status based on the real-time data collected by sensors and the historical data of the vehicle through the LSTM (Long Short-Term Memory Network) model. The data-driven model is flexible and adaptable, but it requires physical mechanism constraints to avoid detection results that exceed the reasonable range.

[0063] It is understandable that the electric vehicle state detection solution provided in the embodiment of the present application will obtain the historical driving data of the vehicle in real time, build and train the data-driven model based on the historical driving data, and build and train the mechanism model based on the physical principle model of the target vehicle. Both the data-driven model and the mechanism model are used for vehicle state detection. Before performing vehicle state detection, the mechanism model has been pre-trained, and the data-driven model has also been pre-trained. In addition, the data model will call the updated historical driving data of the vehicle in real time, and adjust and optimize according to the continuously updated historical driving data.

[0064] Step S130, obtaining a first state detection result output by the data-driven model; and obtaining a second state detection result output by the mechanism model, wherein the first state detection result includes a plurality of first detection values ​​corresponding to a plurality of state indicators of the target vehicle, and the second state detection result includes a plurality of second detection values ​​corresponding to a plurality of state indicators of the target vehicle.

[0065] Among them, the multiple state indicators correspond one-to-one to the multiple first detection values, the multiple state indicators correspond one-to-one to the multiple second detection values, and each state indicator corresponds to one first detection value and one second detection value.

[0066] In a possible embodiment, the data-driven model sets physical boundary conditions and regularization terms to ensure that the output value of the data-driven model complies with physical laws, and the physical boundary conditions indicate that the first state detection result does not exceed multiple preset thresholds of the multiple state indicators; and obtaining the first state detection result output by the data-driven model includes:

[0067] Processing the driving characteristic data through an LSTM model to obtain the first state detection result, wherein the LSTM model is trained based on the historical driving characteristic data of the target vehicle;

[0068] It is detected that the first state detection result meets the physical boundary condition.

[0069] Among them, in order to avoid the data-driven model outputting results that do not conform to the laws of physics, a soft constraint method can be used, that is, guiding and optimizing the output of the data-driven model through physical laws. The soft constraint refers to the use of physical boundary conditions (such as maximum battery current, minimum reaction time, etc.) as constraints in the prediction process of the data-driven model, and punishing the output that violates the laws of physics through a penalty function or regularization term to ensure that the final output results conform to the actual physical limitations.

[0070] Among them, multiple preset thresholds of multiple status indicators may include maximum current, maximum voltage, minimum voltage, charging rate, discharge rate, temperature range, etc. for the battery; and, maximum speed, maximum torque, maximum power, temperature range, etc. for the motor; and, maximum acceleration, maximum deceleration, maximum speed, minimum turning radius, etc. for vehicle dynamics; and, measurement range, accuracy, response time, etc. for the sensor.

[0071] For example, the maximum charging current of the battery is max As a constraint, when the data-driven model predicts the current I predicted When this range is exceeded, the following penalty function is used:

[0072] Penalty=λ*max(0,I predicted -I max )

[0073] Among them, Penalty refers to the penalty value, and λ is the penalty coefficient, which ensures that the output of the data model does not exceed a reasonable range.

[0074] Furthermore, for some physical constraints that cannot be directly hard-limited, the model output is guided by regularization terms, for example, in the prediction of driver reaction time, to ensure that the reaction time does not fall below the physiological limit. Common regularization methods include L1 regularization and L2 regularization.

[0075] It is understandable that by guiding the data-driven model through the soft constraint mechanism, the system can ensure that the detection results conform to the laws of physics, while avoiding the physical distortion problems caused by a single data model and ensuring the robustness of the output results.

[0076] In a possible embodiment, the mechanism model includes a plurality of sub-models, and the sub-models include a dynamic model, a battery charging model, a wear model, and a fatigue model; and obtaining the second state detection result output by the mechanism model includes:

[0077] Obtaining a plurality of initial output values ​​of the plurality of sub-models;

[0078] Determine a plurality of target initial output values ​​corresponding to the plurality of state indicators, wherein the plurality of target initial output values ​​are used to calculate and obtain the detection results corresponding to the plurality of state indicators;

[0079] The second state detection result is determined according to the multiple target initial output values.

[0080] Among them, the target initial value may include battery power, battery charging status, battery health status, battery temperature, number of charge and discharge cycles, etc. for battery status indicators; and motor power, current and voltage, motor capacity conversion efficiency, etc. for power system indicators; and speed, acceleration, mileage, etc. for driving status indicators; and external temperature, road condition information, etc. for environment and external condition indicators; and fault codes, abnormal behaviors, etc. for fault detection indicators; and braking system status, stability control status, etc. for safety indicators; and charging time, cruising range, etc. for user experience indicators; and energy consumption rate, regenerative braking energy recovery rate, etc. for energy consumption indicators.

[0081] It can be seen that in this embodiment, the first state detection result of the electric vehicle state under the simulated complex working condition can be obtained through the data-driven model; and the second state detection result of the electric vehicle state under the simulated simple working condition can be obtained through the mechanism model. Among them, the first state detection result has the problem of inaccurate data, and the second state detection result has the problem of insufficient accuracy due to the inflexibility of the mechanism model.

[0082] Step S140: determining a first weight value of the first state detection result and a second weight value of the second state detection result according to the driving characteristic data and the historical driving characteristic data.

[0083] In a possible embodiment, determining a first weight value of the first state detection result and a second weight value of the second state detection result according to the driving characteristic data and the historical driving characteristic data includes:

[0084] Scoring the driving characteristic data according to the historical driving characteristic data to obtain a data quality score, wherein the data quality score is positively correlated with the data quality of the driving characteristic data;

[0085] Determine the first weight value according to the data quality score, wherein the data quality score is positively correlated with the first weight value;

[0086] The second weight value is determined according to the first weight value.

[0087] Specifically, the first weight value is determined according to the data quality score. Specifically, the first weight value can be obtained by a dynamic weight adjustment formula, and the second weight value is obtained according to the sum of the first weight value and the second weight value being 1. The dynamic weight adjustment formula is as follows:

[0088]

[0089] Among them, Qdata is the data quality score, w data is the first weight value, and θ is the threshold used to control the growth rate of the weight.

[0090] Among them, a larger θ value will make the weight change more smoothly, while a smaller θ value will make the weight more sensitive to changes in data quality scores. In practical applications, the value of θ can be adjusted according to specific circumstances to achieve the best fusion effect.

[0091] In a possible embodiment, the step of performing data scoring on the driving characteristic data according to the historical driving characteristic data to obtain a data quality score includes:

[0092] Determining at least one scoring indicator for the driving characteristic data, wherein the at least one scoring indicator is used to evaluate the data quality of the driving characteristic data from different indicator dimensions;

[0093] Performing data scoring on the driving characteristic data according to the historical driving characteristic data to obtain at least one scoring result of the driving characteristic data, wherein the at least one scoring result corresponds to the at least one scoring indicator in a one-to-one manner;

[0094] The at least one scoring result is weighted and summed according to a preset weight value to obtain the data quality score, wherein the preset weight value includes at least one sub-weight value, and the at least one sub-weight value corresponds one-to-one to the at least one scoring indicator.

[0095] The scoring indicators may include data integrity, data accuracy, data consistency, data relevance, data timeliness and other indicators, and multiple indicators are used to evaluate data quality.

[0096] Among them, the data quality score can be the scoring result of a single indicator, but usually refers to the comprehensive scoring result of multiple scoring indicators. Each scoring indicator corresponds to a preset weight value, and the data quality score can be obtained by weighted summing up each scoring indicator.

[0097] It can be seen that in this embodiment, by obtaining the data quality score of the driving characteristic data, the weights of the mechanism model and the data-driven model can be dynamically adjusted according to the weight dynamic adjustment formula. When the data quality score is high, the weight of the data-driven model will increase, thereby enhancing its flexibility; and when the data quality score is low, the weight of the mechanism model will increase relatively, thereby improving the robustness of the overall prediction.

[0098] Step S150: performing weighted summation on the first state detection result and the second state detection result according to the first weight value and the second weight value to obtain a target state detection result.

[0099] In a possible embodiment, the multiple state indicators correspond one-to-one to the multiple first detection values, the multiple state indicators correspond one-to-one to the multiple second detection values, each of the state indicators corresponds to a first detection value and a second detection value; the first state detection result and the second state detection result are weighted and summed according to the first weight value and the second weight value to obtain a target state detection result, and the method includes:

[0100] Determine the first detection value and the second detection value corresponding to each of the state indicators;

[0101] Performing a weighted summation of the first detection value and the second monitoring value corresponding to each of the state indicators according to the first weight value and the second weight value to obtain a weighted detection value corresponding to each of the state indicators;

[0102] The target state detection result is obtained according to the multiple weighted detection values ​​corresponding to the multiple state indicators.

[0103] Specifically, the target state detection result can be expressed by the following formula:

[0104] Output fusion =w data *Output data +w mechanistic *Output mechanistic

[0105] Among them, Output data Output is the first state detection result output by the data-driven model. mechanistic is the second state detection result output by the mechanism model, w data and w mechanistic are the weight values ​​of the two models, satisfying w data +w mechanistic =1.

[0106] Among them, the target state detection result can not only characterize the current operating state of the vehicle, but also include the prediction results of the vehicle state. The prediction results can be applied to the battery management system of electric vehicles, that is, by accurately predicting the battery life and energy consumption state, optimizing the battery charging and discharging strategy and improving the battery life; and, it can be applied to vehicle health monitoring and predictive maintenance, that is, by predicting the wear of key vehicle components (such as tires, brake pads, etc.), formulating maintenance plans and reducing unexpected failures; and, it can be applied to the intelligent dispatching of electric vehicles, that is, the dispatching system intelligently allocates driving tasks according to the detected vehicle state to ensure driving safety and reduce driving costs; and, it can be applied to energy consumption management, that is, based on the current road conditions, driving behavior and vehicle state, predicting energy consumption and providing optimization suggestions for the driver; and, it can be applied to the intelligent driving assistance system, that is, through the state detection results, helping the driver to understand the vehicle state in real time during driving, optimize driving decisions, and ensure driving safety.

[0107] It can be seen that in this embodiment, by integrating the mechanism model with the data-driven model, based on the dynamic weight switching mechanism, the weight is dynamically adjusted according to the real-time data quality and model error. When the sensor data is complete and stable, the weight of the data-driven model can be increased; when the data is abnormal or the error is large, the weight of the mechanism model is increased. This is conducive to ensuring that reliable state prediction can be maintained in complex situations.

[0108] See also Figure 2 , Figure 2 is a system architecture diagram of an electric vehicle status detection system provided in an embodiment of the present application, such as Figure 2 As shown, the electric vehicle state detection system includes a data-driven model, a mechanism model and a fusion detection model.

[0109] Among them, the data-driven model mainly obtains historical data and current driving data, creates and trains the LSTM model through historical data, preprocesses and extracts features of the current driving data and then inputs it into the LSTM model. The driving data is processed by the LSTM model, and data detection is performed based on the state indicators of the target vehicle state detection to obtain the first state detection result.

[0110] Among them, the mechanism model mainly obtains the current driving data, pre-processes and extracts the features of the current driving data, and then inputs it into multiple models that have been trained in advance based on the physical principles of the target vehicle. The multiple models include dynamic models, charging and discharging models, wear models, and fatigue models, etc. Then, the mechanism detection is performed according to the state indicators of the target vehicle state detection this time to obtain the second state detection result.

[0111] The fusion detection model receives the first state detection result and the second state detection result, performs data preprocessing and feature extraction on the detection results, assigns weight values ​​to the first state detection result and the second state detection result based on the adaptive weight, and performs weighted summation to obtain the target state detection result. The adaptive weight is presented in an adaptive weight formula based on the data quality score.

[0112] It can be seen that in this embodiment, data-driven detection is performed by combining historical data with current driving data and the LSTM model, and mechanism model detection is performed based on physical laws. Finally, through the adaptive weighting mechanism, the system fuses the outputs of the two models according to data quality and model error to generate detection results of key status indicators. This is conducive to improving the accuracy and stability of vehicle status detection.

[0113] See also Figure 3 , Figure 3 is an overall flow chart of an electric vehicle state detection method provided in an embodiment of the present application, such as Figure 3 As shown, the electric vehicle state detection method comprises the following steps:

[0114] Step S310, creating and training a data-driven model based on historical driving characteristic data of the target vehicle; and creating and training a mechanism model based on the physical principles of the target vehicle.

[0115] The data-driven model mainly detects the health status based on the real-time data collected by sensors and the historical data of the vehicle through the LSTM (Long Short-Term Memory Network) model. The data-driven model is flexible and adaptable, but it requires physical mechanism constraints to avoid detection results that exceed the reasonable range.

[0116] Among them, the mechanism model is mainly based on the physical laws of the vehicle (such as dynamic model, battery model, driver fatigue model), which can provide accurate predictions under steady-state conditions. However, since the mechanism model is difficult to adapt to complex working conditions and dynamic changes, the model accuracy is limited in practical applications.

[0117] Step S320, obtaining the current driving characteristic data of the target vehicle.

[0118] In a possible embodiment, the step of obtaining the current driving characteristic data of the target vehicle includes:

[0119] Acquire current driving data of the target vehicle;

[0120] Acquire the multiple status indicators for status detection of the target vehicle, where the status indicators are used to detect the status of the target vehicle from multiple indicator dimensions;

[0121] Determine target driving data corresponding to the multiple state indicators, wherein the target driving data is used to obtain detection results corresponding to the multiple state indicators;

[0122] A target operation is performed on the target driving data to obtain the driving feature data, wherein the target operation includes a preprocessing operation and a feature engineering operation.

[0123] Among them, the status indicators of electric vehicles usually include at least one of multiple indicators such as battery status indicators, power system indicators, driving status indicators, environmental and external conditions indicators, fault detection indicators, safety indicators, user experience indicators and energy consumption indicators. The status indicators can be set exclusively according to user needs and the vehicle's own status, and different vehicles can use different status indicators.

[0124] Step S330, inputting the driving characteristic data into the data driven model and the mechanism model respectively.

[0125] Step S340, obtaining a first state detection result output by the data-driven model; and obtaining a second state detection result output by the mechanism model.

[0126] Among them, the first state detection result includes multiple first detection values ​​corresponding to multiple state indicators of the target vehicle, and the second state detection result includes multiple second detection values ​​corresponding to multiple state indicators of the target vehicle. The multiple state indicators correspond one-to-one to the multiple first detection values, and the multiple state indicators correspond one-to-one to the multiple second detection values, and each state indicator corresponds to a first detection value and a second detection value.

[0127] Step S350: whether the first state detection result meets the physical boundary condition.

[0128] If it is detected that the first state detection result meets the physical boundary condition, step S360 is executed; and if it is detected that the first state detection result does not meet the physical boundary condition, step S370 is executed.

[0129] The physical boundary condition indicates that the first state detection result does not exceed the multiple preset thresholds of the multiple state indicators. The multiple preset thresholds of the multiple state indicators may include the maximum current, maximum voltage, minimum voltage, charging rate, discharging rate, temperature range, etc. for the battery; and the maximum speed, maximum torque, maximum power, temperature range, etc. for the motor; and the maximum acceleration, maximum deceleration, maximum speed, minimum turning radius, etc. for vehicle dynamics; and the measurement range, accuracy, response time, etc. for the sensor.

[0130] Step S360: determining the data quality score of the driving characteristic data.

[0131] In a possible embodiment, determining the data quality score of the driving characteristic data includes:

[0132] Determining at least one scoring indicator for the driving characteristic data, wherein the at least one scoring indicator is used to evaluate the data quality of the driving characteristic data from different indicator dimensions;

[0133] Performing data scoring on the driving characteristic data according to the historical driving characteristic data to obtain at least one scoring result of the driving characteristic data, wherein the at least one scoring result corresponds one-to-one to the at least one scoring indicator;

[0134] The at least one scoring result is weighted and summed according to a preset weight value to obtain the data quality score, wherein the preset weight value includes at least one sub-weight value, and the at least one sub-weight value corresponds one-to-one to the at least one scoring indicator.

[0135] The scoring indicators may include data integrity, data accuracy, data consistency, data relevance, data timeliness and other indicators, and multiple indicators are used to evaluate data quality.

[0136] Step S370: adjusting the data driven model according to the first state detection result.

[0137] Step S380: Determine a first weight value of the first state detection result and a second weight value of the second state detection result according to the data quality score.

[0138] Specifically, the first weight value is determined according to the data quality score. Specifically, the first weight value can be obtained by a dynamic weight adjustment formula, and the second weight value is obtained according to the sum of the first weight value and the second weight value being 1. The dynamic weight adjustment formula is as follows:

[0139]

[0140] Among them, Q data is the data quality score, w data is the first weight value, and θ is the threshold used to control the growth rate of the weight.

[0141] Step S390, performing weighted summation on the first state detection result and the second state detection result according to the first weight value and the second weight value to obtain a target state detection result.

[0142] Specifically, the target state detection result can be expressed by the following formula:

[0143] Output fusion =w data *Output data +w mechanistic *Outputmechanistic

[0144] Among them, Output data Output is the first state detection result output by the data-driven model. mechanistic is the second state detection result output by the mechanism model, w data and w mechanistic are the weight values ​​of the two models, satisfying w data +w mechanistic =1.

[0145] It can be seen that in this embodiment, by fusing the data-driven model and the mechanism model, the weights of the mechanism model and the data-driven model are dynamically adjusted according to the dynamic weight switching mechanism, so that when the data is complete and stable, the weight of the data-driven model is increased; when the data is abnormal or the error is large, the weight of the mechanism model is increased, which can maintain reliable vehicle status detection in complex situations, and is conducive to improving the accuracy and robustness of vehicle status detection.

[0146] See also Figure 4 , Figure 4 A functional unit block diagram of an electric vehicle state detection device 400 provided in an embodiment of the present application is shown in FIG. Figure 4 As shown, the electric vehicle state detection device 400 includes the following units:

[0147] A first acquisition unit 410 is used to acquire current driving characteristic data of the target vehicle;

[0148] An input unit 420, used to input the driving characteristic data into a pre-trained data-driven model and a mechanism model, respectively, wherein the data-driven model is trained based on the historical driving characteristic data of the target vehicle, and the mechanism model is trained based on the physical principle of the target vehicle;

[0149] A second acquisition unit 430 is used to acquire a first state detection result output by the data-driven model; and acquire a second state detection result output by the mechanism model, wherein the first state detection result includes a plurality of first detection values ​​corresponding to a plurality of state indicators of the target vehicle, and the second state detection result includes a plurality of second detection values ​​corresponding to a plurality of state indicators of the target vehicle;

[0150] A determination unit 440, configured to determine a first weight value of the first state detection result and a second weight value of the second state detection result according to the driving characteristic data and the historical driving characteristic data;

[0151] The state detection unit 450 is used to perform weighted summation on the first state detection result and the second state detection result according to the first weight value and the second weight value to obtain a target state detection result.

[0152] In one embodiment, determining a first weight value of the first state detection result and a second weight value of the second state detection result according to the driving characteristic data and the historical driving characteristic data includes:

[0153] Scoring the driving characteristic data according to the historical driving characteristic data to obtain a data quality score, wherein the data quality score is positively correlated with the data quality of the driving characteristic data;

[0154] Determine the first weight value according to the data quality score, wherein the data quality score is positively correlated with the first weight value;

[0155] The second weight value is determined according to the first weight value.

[0156] In one embodiment, the step of performing data scoring on the driving characteristic data according to the historical driving characteristic data to obtain a data quality score includes:

[0157] Determining at least one scoring indicator for the driving characteristic data, wherein the at least one scoring indicator is used to evaluate the data quality of the driving characteristic data from different indicator dimensions;

[0158] Performing data scoring on the driving characteristic data according to the historical driving characteristic data to obtain at least one scoring result of the driving characteristic data, wherein the at least one scoring result corresponds to the at least one scoring indicator in a one-to-one manner;

[0159] The at least one scoring result is weighted and summed according to a preset weight value to obtain the data quality score, wherein the preset weight value includes at least one sub-weight value, and the at least one sub-weight value corresponds one-to-one to the at least one scoring indicator.

[0160] In one embodiment, the step of obtaining the current driving characteristic data of the target vehicle includes:

[0161] Acquire current driving data of the target vehicle;

[0162] Acquire the multiple status indicators for status detection of the target vehicle, where the status indicators are used to detect the status of the target vehicle from multiple indicator dimensions;

[0163] Determine target driving data corresponding to the multiple state indicators, wherein the target driving data is used to obtain detection results corresponding to the multiple state indicators;

[0164] A target operation is performed on the target driving data to obtain the driving feature data, wherein the target operation includes a preprocessing operation and a feature engineering operation.

[0165] In one embodiment, the multiple state indicators correspond one-to-one to the multiple first detection values, the multiple state indicators correspond one-to-one to the multiple second detection values, each of the state indicators corresponds to a first detection value and a second detection value; the first state detection result and the second state detection result are weighted and summed according to the first weight value and the second weight value to obtain a target state detection result, and the method includes:

[0166] Determine the first detection value and the second detection value corresponding to each of the state indicators;

[0167] Performing a weighted summation of the first detection value and the second monitoring value corresponding to each of the state indicators according to the first weight value and the second weight value to obtain a weighted detection value corresponding to each of the state indicators;

[0168] The target state detection result is obtained according to the multiple weighted detection values ​​corresponding to the multiple state indicators.

[0169] In one embodiment, the data-driven model sets physical boundary conditions and regularization terms to ensure that the output value of the data-driven model complies with physical laws, and the physical boundary conditions indicate that the first state detection result does not exceed multiple preset thresholds of the multiple state indicators; the obtaining of the first state detection result output by the data-driven model includes:

[0170] Processing the driving characteristic data through an LSTM model to obtain the first state detection result, wherein the LSTM model is trained based on the historical driving characteristic data of the target vehicle;

[0171] It is detected that the first state detection result meets the physical boundary condition.

[0172] In one embodiment, the mechanism model includes a plurality of sub-models, including a kinetic model, a battery charging model, a wear model, and a fatigue model; and obtaining the second state detection result output by the mechanism model includes:

[0173] Obtaining a plurality of initial output values ​​of the plurality of sub-models;

[0174] Determine a plurality of target initial output values ​​corresponding to the plurality of state indicators, wherein the plurality of target initial output values ​​are used to calculate and obtain the detection results corresponding to the plurality of state indicators;

[0175] The second state detection result is determined according to the multiple target initial output values.

[0176] It can be understood that since the method embodiment and the device embodiment are different presentation forms of the same technical concept, the content of the method embodiment part in this application should be synchronously adapted to the device embodiment part and will not be repeated here.

[0177] It can be seen that the device first obtains the current driving characteristic data of the target vehicle through the first acquisition unit 410; then, the driving characteristic data is input into the pre-trained data-driven model and the mechanism model through the input unit 420; then, the first state detection result output by the data-driven model is obtained through the second acquisition unit 430; and the second state detection result output by the mechanism model is obtained; secondly, the first weight value of the first state detection result and the second weight value of the second state detection result are determined by the determination unit 440 according to the driving characteristic data and the historical driving characteristic data; finally, the first state detection result and the second state detection result are weighted and summed according to the first weight value and the second weight value by the state detection unit 450 to obtain the target state detection result. The weights of the mechanism model and the data-driven model are dynamically adjusted. When the data is complete and stable, the weight of the data-driven model is increased; when the data is abnormal or the error is large, the weight of the mechanism model is increased, the advantages and disadvantages of the two models are balanced, and the flexibility and robustness of the vehicle state detection are improved.

[0178] See also Figure 5 , Figure 5 is a structural block diagram of an electronic device provided in an embodiment of the present application, such as Figure 5 As shown, the electronic device 500 may include one or more of the following components: a processor 510, and a memory 520 coupled to the processor 510, wherein the memory 520 may store one or more computer programs, and the one or more computer programs may be configured to implement the methods described in the above embodiments when executed by one or more processors 510.

[0179] The processor 510 may include one or more processing cores. The processor 510 uses various interfaces and lines to connect various parts of the entire electronic device 500, and executes various functions and processes data of the electronic device 500 by running or executing instructions, programs, code sets or instruction sets stored in the memory 520, and calling data stored in the memory 520. Optionally, the processor 510 can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 510 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 510, but may be implemented separately through a communication chip.

[0180] The memory 520 may include a random access memory (RAM) or a read-only memory (ROM). The memory 520 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 520 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc. The data storage area may also store data created by the electronic device 500 during use, etc.

[0181] It is understandable that the electronic device 500 may include more or fewer structural elements than those in the above structural block diagram, for example, including a power module, physical buttons, a Wi-Fi module, a speaker, a Bluetooth module, a sensor, etc., which are not limited here.

[0182] In addition, an embodiment of the present application also provides a computer storage medium, which stores a computer program that can be loaded by a processor and executes the above-mentioned electric vehicle status detection method. The computer-readable storage medium includes, for example: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0183] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described action sequence, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and units involved are not necessarily required by the present application.

[0184] This is only a logical function division, and there may be other divisions in actual implementation; for example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, which may be electrical, mechanical or other forms.

[0185] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0186] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may be physically included separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0187] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform some steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a volatile memory or a non-volatile memory. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM) or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (double data rate SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (enhanced SDRAM, ESDRAM), synchronous link dynamic random access memory (synchlink DRAM, SLDRAM) and direct memory bus random access memory (directrambus RAM, DR RAM) and other media that can store program code.

[0188] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0189] The embodiments of the present application are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application. At the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

[0190] Although the present application is disclosed as above, the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions without departing from the spirit and scope of the present application, and can make various changes and modifications, including the combination of the above-mentioned different functions and implementation steps, including software and hardware implementation methods, all of which are within the scope of protection of the present application.

Claims

1. A method for detecting the state of an electric vehicle, characterized in that: include: Acquiring current driving characteristic data of the target vehicle, wherein the driving characteristic data is used to characterize driving parameters of the target vehicle; Inputting the driving characteristic data into a pre-trained data-driven model and a mechanism model respectively, wherein the data-driven model is trained based on the historical driving characteristic data of the target vehicle, and the mechanism model is trained based on the physical principle of the target vehicle; Obtaining a first state detection result output by the data-driven model; and obtaining a second state detection result output by the mechanism model, wherein the first state detection result includes a plurality of first detection values ​​corresponding to a plurality of state indicators of the target vehicle, and the second state detection result includes a plurality of second detection values ​​corresponding to a plurality of state indicators of the target vehicle; Determine a first weight value of the first state detection result and a second weight value of the second state detection result according to the driving characteristic data and the historical driving characteristic data; The first state detection result and the second state detection result are weightedly summed according to the first weight value and the second weight value to obtain a target state detection result.

2. The method according to claim 1, characterized in that: The determining, according to the driving characteristic data and the historical driving characteristic data, a first weight value of the first state detection result and a second weight value of the second state detection result comprises: Scoring the driving characteristic data according to the historical driving characteristic data to obtain a data quality score, wherein the data quality score is positively correlated with the data quality of the driving characteristic data; Determine the first weight value according to the data quality score, wherein the data quality score is positively correlated with the first weight value; The second weight value is determined according to the first weight value.

3. The method according to claim 2, characterized in that The step of performing data scoring on the driving characteristic data according to the historical driving characteristic data to obtain a data quality score includes: Determining at least one scoring indicator for the driving characteristic data, wherein the at least one scoring indicator is used to evaluate the data quality of the driving characteristic data from different indicator dimensions; Performing data scoring on the driving characteristic data according to the historical driving characteristic data to obtain at least one scoring result of the driving characteristic data, wherein the at least one scoring result corresponds to the at least one scoring indicator in a one-to-one manner; The at least one scoring result is weighted and summed according to a preset weight value to obtain the data quality score, wherein the preset weight value includes at least one sub-weight value, and the at least one sub-weight value corresponds one-to-one to the at least one scoring indicator.

4. The method according to claim 3, characterized in that The step of obtaining the current driving characteristic data of the target vehicle includes: Acquire current driving data of the target vehicle; Acquire the multiple status indicators for status detection of the target vehicle, where the status indicators are used to detect the status of the target vehicle from multiple indicator dimensions; Determine target driving data corresponding to the multiple state indicators, wherein the target driving data is used to obtain detection results corresponding to the multiple state indicators; A target operation is performed on the target driving data to obtain the driving feature data, wherein the target operation includes a preprocessing operation and a feature engineering operation.

5. The method according to any one of claims 1 to 4, characterized in that: The multiple state indicators correspond one-to-one to the multiple first detection values, the multiple state indicators correspond one-to-one to the multiple second detection values, and each of the state indicators corresponds to a first detection value and a second detection value; The first state detection result and the second state detection result are weighted and summed according to the first weight value and the second weight value to obtain a target state detection result, the method comprising: Determine the first detection value and the second detection value corresponding to each of the state indicators; Performing a weighted summation of the first detection value and the second monitoring value corresponding to each of the state indicators according to the first weight value and the second weight value to obtain a weighted detection value corresponding to each of the state indicators; The target state detection result is obtained according to the multiple weighted detection values ​​corresponding to the multiple state indicators.

6. The method according to claim 5, characterized in that In the data-driven model, physical boundary conditions and regularization terms are set to ensure that the output value of the data-driven model complies with physical laws, wherein the physical boundary conditions indicate that the first state detection result does not exceed multiple preset thresholds of the multiple state indicators; The obtaining of the first state detection result output by the data-driven model includes: Processing the driving characteristic data through an LSTM model to obtain the first state detection result, wherein the LSTM model is trained based on the historical driving characteristic data of the target vehicle; It is detected that the first state detection result meets the physical boundary condition.

7. The method according to claim 6, characterized in that The mechanism model includes a plurality of sub-models, including a dynamics model, a battery charging model, a wear model and a fatigue model; the second state detection result output by the mechanism model is obtained, including: Obtaining a plurality of initial output values ​​of the plurality of sub-models; Determine a plurality of target initial output values ​​corresponding to the plurality of state indicators, wherein the plurality of target initial output values ​​are used to calculate and obtain the detection results corresponding to the plurality of state indicators; The second state detection result is determined according to the multiple target initial output values.

8. An electric vehicle state detection device, characterized in that: The device comprises: A first acquisition unit, used to acquire current driving characteristic data of the target vehicle; An input unit, used to input the driving characteristic data into a pre-trained data-driven model and a mechanism model, respectively, wherein the data-driven model is trained based on the historical driving characteristic data of the target vehicle, and the mechanism model is trained based on the physical principle of the target vehicle; A second acquisition unit is used to acquire a first state detection result output by the data-driven model; and acquire a second state detection result output by the mechanism model, wherein the first state detection result includes a plurality of first detection values ​​corresponding to a plurality of state indicators of the target vehicle, and the second state detection result includes a plurality of second detection values ​​corresponding to a plurality of state indicators of the target vehicle; a determining unit, configured to determine a first weight value of the first state detection result and a second weight value of the second state detection result according to the driving characteristic data and the historical driving characteristic data; A state detection unit is used to perform weighted summation on the first state detection result and the second state detection result according to the first weight value and the second weight value to obtain a target state detection result.

9. An electronic device, characterized in that: The method comprises a processor, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the programs include instructions for executing the steps in the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: The computer program / instructions are executed by a processor to implement the steps of the method according to any one of claims 1 to 7.

Images