Pre-tightening control method and device of electric vehicle, electric vehicle and electronic equipment

Abstract

The invention provides a pre-tightening control method and device for an electric vehicle, the electric vehicle, electronic equipment and a storage medium, and the method comprises the steps that acoustic signals of friction between tires of the electric vehicle and the ground are obtained; key feature extraction is carried out on the preprocessed acoustic signals to obtain key features; performing emergency detection according to the key features to obtain a detection result; if the detection result is that an emergency occurs, road adhesion prediction is carried out according to the key features, and a predicted friction coefficient is obtained; and performing pre-tightening control on the electric vehicle according to the predicted friction coefficient. By implementing the application, the pre-tightening control method can be given in time according to the captured road surface change condition, the interference of the abnormal road surface on driving is reduced, the response is faster, the control on the electric vehicle is more stable and effective, the dependence on hardware is less, and various driving scenes can be met.

Description

Technical Field

[0001] The present application relates to the technical field of electric vehicle control. Specifically, it relates to a pre-tightening control method, device, electric vehicle, electronic device, and storage medium for an electric vehicle. Background Art

[0002] Perceiving the road surface through sound or acoustic signals is a common method for pre-tightening control of electric vehicles. In the prior art, there are various ways to perceive and pre-tighten the road surface through acoustic signals, such as classification methods based on tire noise spectrum analysis, multi-modal schemes combining vibration and acoustic sensors, forward-looking systems integrating lidar and microphones, indirect control schemes relying on rain sensors, and prediction models using deep learning of raw audio signals, etc.

[0003] However, these above-mentioned schemes generally have defects such as response delay, multi-sensor dependence, single control strategy, or poor environmental robustness. They cannot properly handle the interference of abnormal road surfaces, are easily affected by the surrounding environment, resulting in poor pre-tightening control ability for electric vehicles, higher implementation costs, and unable to meet various usage requirements. Summary of the Invention

[0004] The purpose of the present application is to provide a pre-tightening control method, device, electric vehicle, electronic device, and storage medium for an electric vehicle, which can give a pre-tightening control method in a timely manner according to the captured changes in the road surface, reduce the interference of abnormal road surfaces on driving, have a faster response, more stable and effective control of the electric vehicle, less dependence on hardware, and be able to meet various driving scenarios.

[0005] In a first aspect, an embodiment of the present application provides a pre-tightening control method for an electric vehicle, the method comprising:

[0006] A pre-tightening control method for an electric vehicle, characterized in that the method comprises:

[0007] Obtain the acoustic signal of the friction between the electric vehicle tire and the ground;

[0008] Extract key features from the preprocessed acoustic signal to obtain key features;

[0009] Perform emergency event detection based on the key features to obtain a detection result;

[0010] If the detection result is that an emergency event is encountered, perform road surface adhesion prediction based on the key features to obtain a predicted friction coefficient;

[0011] Perform pre-tightening control on the electric vehicle according to the predicted friction coefficient.

[0012] In the above implementation process, the detection of emergencies and the prediction of road adhesion are carried out in sequence according to the key features. The electric vehicle is pre-tightened according to the predicted friction coefficient. The pre-tightening control method can be given in time according to the captured road surface changes, reducing the interference of abnormal road surfaces on driving, with faster response, more stable and effective control of the electric vehicle, less dependence on hardware, and can meet a variety of driving scenarios.

[0013] Further, the step of extracting key features from the preprocessed acoustic signal to obtain key features includes:

[0014] Segment the preprocessed acoustic signal to obtain multiple segmented sound signals;

[0015] Analyze each of the multiple segmented sound signals to obtain sound intensities at multiple different frequencies;

[0016] Determine the ratio of the sound intensity in the first threshold frequency range to the sound intensities at multiple different frequencies as the high-frequency energy ratio;

[0017] Detect the change in the resonance frequency of the tire in the preprocessed acoustic signal according to the second threshold frequency range to obtain the resonance peak offset rate;

[0018] Extract the transient impact count of the tire from the preprocessed acoustic signal according to the third threshold;

[0019] Determine the high-frequency energy ratio, the resonance peak offset rate, and the transient impact count as the key features.

[0020] In the above implementation process, different key features are extracted from the acoustic signal respectively to obtain the acoustic performance of the acoustic signal under different conditions, which can truly reflect the road surface changes during the driving of the electric vehicle and provide data support for pre-tightening control.

[0021] Further, the step of detecting emergencies according to the key features to obtain a detection result includes:

[0022] Input the key features into a pre-trained emergency detection model for detection to obtain a mutation probability value;

[0023] Judge whether the mutation probability value is greater than the fourth threshold;

[0024] If so, determine the detection result as encountering an emergency.

[0025] In the above implementation process, detecting emergencies for key features can detect whether there are mutations on the road surface during driving, reduce the impact of abnormal road surfaces on the driving of electric vehicles, improve the robustness of the vehicle, and provide safety guarantees for the pre-tightening control of electric vehicles.

[0026] Further, the step of predicting road surface adhesion according to the key features to obtain the predicted friction coefficient includes:

[0027] Obtain the sound characteristics of different types of road surfaces and their corresponding actual friction coefficients;

[0028] Construct a mapping relationship between the sound characteristics of different types of road surfaces and the corresponding actual friction coefficients;

[0029] When the detection result is encountering an emergency, input the key features into a pre-constructed road surface adhesion prediction model to obtain the predicted friction coefficient.

[0030] In the above implementation process, the basic situation of the road surface is obtained in advance according to the mapping relationship between the sound characteristics of different road surfaces and their actual friction coefficients, which is convenient for accurately predicting the friction coefficient, reducing the data error generated in the prediction process, and reducing the interference of abnormal road surface conditions.

[0031] Further, the step of performing pre-tightening control on the electric vehicle according to the predicted friction coefficient includes:

[0032] Obtain the torque limit data corresponding to the predicted friction coefficient;

[0033] When the torque limit data is less than the fifth threshold, perform cooperative pre-boosting on the electric vehicle and control the torque rising speed to achieve pre-tightening control of the electric vehicle.

[0034] In the above implementation process, cooperative pre-boosting is performed on the electric vehicle according to the torque limit data, so that the electric vehicle can pressurize the brake caliper in advance and give a quick response when braking is required, reducing the safety hazard brought by torque limitation.

[0035] Further, after the step of performing pre-tightening control on the electric vehicle according to the predicted friction coefficient, it further includes:

[0036] Obtain the actual slip rate of the electric vehicle;

[0037] Perform error correction on the predicted friction coefficient according to the actual slip rate to obtain an error rate;

[0038] Adjust the parameters of the road surface adhesion prediction model according to the error rate.

[0039] In the above implementation process, the actual friction data during the driving of the electric vehicle is deduced based on the actual slip rate of the electric vehicle, and the predicted friction data is corrected for errors according to the actual friction data, and the parameter settings of the road surface adhesion prediction model are adjusted, which can improve the accuracy of the road surface adhesion prediction model and the accuracy of the predicted friction coefficient.

[0040] In a second aspect, an embodiment of the present application further provides a pre-tightening control device for an electric vehicle, and the device includes:

[0041] An acquisition module, configured to acquire an acoustic signal of the friction between the electric vehicle tire and the ground;

[0042] An extraction module, configured to extract key features from the preprocessed acoustic signal to obtain key features;

[0043] A detection module, configured to perform an emergency detection according to the key features to obtain a detection result;

[0044] A prediction module, configured to, if the detection result is that an emergency is encountered, perform a road surface adhesion prediction according to the key features to obtain a predicted friction coefficient;

[0045] A pre-tightening control module, configured to perform a pre-tightening control on the electric vehicle according to the predicted friction coefficient.

[0046] In the above implementation process, an emergency detection and a road surface adhesion prediction are sequentially performed according to the key features, and a pre-tightening control is performed on the electric vehicle according to the predicted friction coefficient. A pre-tightening control method can be given in a timely manner according to the captured road surface changes, reducing the interference of abnormal road surfaces on driving, with faster response, more stable and effective control of the electric vehicle, less dependence on hardware, and being able to meet a variety of driving scenarios.

[0047] In a third aspect, an electric vehicle provided by an embodiment of the present application includes the pre-tightening control device for an electric vehicle in the second aspect.

[0048] In a fourth aspect, an electronic device provided by an embodiment of the present application includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the steps of the method according to any one of the first aspect are implemented.

[0049] In a fifth aspect, a computer-readable storage medium provided by an embodiment of the present application has instructions stored thereon, and when the instructions are run on a computer, the computer is caused to execute the method according to any one of the first aspect.

[0050] Other features and advantages of the present disclosure will be described in the subsequent description, or, some features and advantages can be inferred from the description or be undoubtedly determined, or can be learned by implementing the above technologies of the present disclosure.

[0051] And it can be implemented according to the content of the description. The following will be described in detail with reference to the preferred embodiments of the present application and the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] To more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the accompanying drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and should not be regarded as limiting the scope value. For those of ordinary skill in the art, without creative efforts, other relevant drawings can also be obtained based on these drawings.

[0053] Figure 1 It is a schematic flow chart of the pre-tightening control method for an electric vehicle provided by an embodiment of the present application;

[0054] Figure 2 It is a schematic structural composition diagram of the pre-tightening control device for an electric vehicle provided by an embodiment of the present application;

[0055] Figure 3 It is a schematic structural composition diagram of the electronic device provided by an embodiment of the present application. Specific Embodiments

[0056] The following will describe the technical solutions in the embodiments of the present application in conjunction with the accompanying drawings in the embodiments of the present application.

[0057] It should be noted that similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present application, terms such as "first" and "second" are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0058] The following will further describe in detail the specific embodiments of the present application in conjunction with the drawings and embodiments. The following embodiments are used to illustrate the present application but are not used to limit the scope value of the present application.

[0059] In the prior art, most of the methods for an electric vehicle to perform pre-tightening control by using sound or acoustic signals have many problems. For example, using a support vector machine to identify wet and dry roads has a response lag, the multi-modal scheme combining vibration and acoustic sensors relies on hardware redundancy and lacks the ability to predict mutations, the forward-looking system integrating lidar and microphones is difficult to mass-produce due to high costs and environmental sensitivity of optical devices, and the indirect control scheme relying on a rain sensor is easily interfered by local road anomalies and causes false triggering. The above-mentioned schemes have defects such as response delay, excessive hardware dependence, single control strategy or poor robustness.

[0060] In view of the defects existing in the above-mentioned prior art, the present application achieves significant breakthroughs in terms of hardware cost, response speed, and control stability through the collaborative control of forward-looking timing prediction of pure acoustic signals, pre-tightening torque gradient limitation, and the Electronic Stability Program (ESP), as well as a closed-loop self-learning mechanism, thereby solving the pain points of the prior art such as dependence on optical devices, unreliability of black-box models, and insufficient detection of local road surface mutations.

[0061] Embodiment 1

[0062] Figure 1 is a schematic flowchart of a pre-tightening control method for an electric vehicle provided by an embodiment of the present application. As Figure 1 shown, the method includes:

[0063] S1. Obtain the acoustic signal of the friction between the electric vehicle tire and the ground;

[0064] S2. Extract key features from the preprocessed acoustic signal to obtain key features;

[0065] S3. Detect emergencies based on the key features to obtain a detection result;

[0066] S4. If the detection result indicates an emergency, predict the road surface adhesion based on the key features to obtain a predicted friction coefficient;

[0067] S5. Perform pre-tightening control on the electric vehicle according to the predicted friction coefficient.

[0068] In the above implementation process, emergencies are detected and road surface adhesion is predicted in sequence based on the key features, and pre-tightening control is performed on the electric vehicle according to the predicted friction coefficient. A pre-tightening control method can be given in a timely manner according to the captured road surface changes, reducing the interference of abnormal road surfaces on driving, with faster response, more stable and effective control of the electric vehicle, less dependence on hardware, and the ability to meet various driving scenarios.

[0069] This application proposes a method for forward prediction and pre-tightening torque control optimization of electric vehicles based on acoustic signals. By deploying a waterproof directional microphone array on the vehicle chassis, the noise generated by the tire-road contact is collected in real time. After adaptive filtering, the normalized least mean square (NLMS) algorithm is adopted in the embodiments of this application to eliminate environmental interference. Then, the short-time Fourier transform is used to extract the high-frequency sound pressure energy density. By combining the continuous wavelet transform to analyze the instantaneous frequency shift of the tire cavity resonance peak, a multi-dimensional feature vector including the band energy ratio, resonance peak shift rate, and transient impact count is constructed. The long short-term memory network is used to perform temporal modeling on the acoustic features within a 300-ms time window to identify mutation events such as sudden drops in high-frequency energy and spectral dispersion, triggering the Gaussian process regression model to predict the road adhesion friction coefficient within the next 100 - 300 ms. According to the predicted friction coefficient, pre-tightening control commands are generated, including restricting the drive torque gradient and electronically stabilizing the program hydraulic pre-boost. The motor output torque and braking pressure are dynamically adjusted through the CAN bus to constrain the tire slip rate in advance to the safety threshold. At the same time, based on the actual slip rate data fed back by the wheel speed sensor and the inertial measurement unit, the online parameter correction of the acoustic prediction model is carried out to form a full-link control architecture of "acoustic perception, feature analysis, prediction decision-making, pre-tightening execution, and closed-loop verification".

[0070] This application does not require the use of lidar or cameras, and the pre-tightening control ability of electric vehicles can be improved. The smoothness of vehicle handling under abnormal road conditions can be significantly enhanced by restricting the pre-tightening torque gradient.

[0071] In S1, the acoustic signal of the electric vehicle tire-ground friction is obtained by capturing the unique sound of the tire-road friction. After obtaining the acoustic signal, preprocessing is required to filter out irrelevant noise.

[0072] In this application, a waterproof directional microphone is installed at each position on the vehicle chassis close to the four tires, pointing to the tire-ground contact area, for collecting the high-frequency noise generated when the tire rolls.

[0073] The preprocessing process includes noise reduction for the collected tire noise, engine noise, wind noise, etc. By using the adaptive filtering algorithm, the engine speed and vehicle speed information are analyzed in real time to generate a signal with a waveform opposite to that of the interference sound, canceling the noise and retaining the "rustling sound" or "splash sound" of the tire-ground contact.

[0074] The preprocessing process also includes amplifying the noise-reduced acoustic signal and intercepting the key frequency band (0.5 - 8 kHz), removing irrelevant high and low-frequency clutter.

[0075] Further, S2 includes:

[0076] Segment the preprocessed acoustic signal to obtain multiple segmented sound signals;

[0077] Analyze the multiple segmented sound signals respectively to obtain the sound intensities at multiple different frequencies;

[0078] Determine that the proportion of the sound intensity in the first threshold frequency range among the sound intensities at multiple different frequencies is the high-frequency energy ratio;

[0079] Detect the change in the resonance frequency of the tire in the preprocessed acoustic signal according to the second threshold frequency range to obtain the resonance peak shift rate;

[0080] Extract the transient impact count of the tire from the preprocessed acoustic signal according to the third threshold;

[0081] Determine that the high-frequency energy ratio, the resonance peak shift rate, and the transient impact count are key features.

[0082] In the above implementation process, extract different key features from the acoustic signal respectively to obtain the acoustic performance of the acoustic signal under different conditions, which can truly reflect the road surface changes during the driving process of the electric vehicle and provide data support for pre-tightening control.

[0083] In this application, extracting key features is divided into three steps. First, perform short-time Fourier transform, segment the sound signal into 10-millisecond segments, and analyze the sound intensities at different frequencies within each segment. For example: the high-frequency noise (similar to a "hissing" sound) on a dry road surface has high energy, while the high-frequency sound on a wet and slippery road surface (such as a water film) will be weakened and the energy will drop suddenly.

[0084] The high-frequency energy ratio is the proportion of the intensity of the sound in the first threshold frequency range (2 - 4 kHz). Since the sound intensity will decrease when passing through an abnormal road surface (such as a wet and slippery road surface), therefore, calculating the high-frequency energy ratio in this application can indicate passing through an abnormal road surface.

[0085] Second, wavelet transform, detect the resonance frequency of the tire cavity, that is, the "buzzing" sound emitted during the tire impact, that is, the change occurring in the second threshold frequency range (frequency about 200 - 800 Hz). When the tire presses on an abnormal road surface condition (such as a wet and slippery road surface), the resonance frequency will suddenly shift.

[0086] The resonance peak shift rate is the change speed of the resonance frequency (a sudden shift represents a sudden change in road surface adhesion).

[0087] The transient impact count is the number of "pop" sounds when the tire presses over water or gravel. Select the acoustic signal that conforms to the "pop" sound when the tire presses over water or gravel according to the third threshold.

[0088] Further, S3 includes:

[0089] Input the key features into a pre-trained emergency event detection model for detection to obtain a mutation probability value;

[0090] Determine whether the mutation probability value is greater than the fourth threshold;

[0091] If so, determine that the detection result is encountering an emergency event.

[0092] In the above implementation process, detecting emergency events for key features can detect whether there are mutations on the road surface during the driving process, reduce the impact of abnormal road surfaces on the driving of electric vehicles, improve the robustness of the vehicle, and provide safety guarantees for the pre-tightening control of electric vehicles.

[0093] The purpose of detecting emergency events is to determine whether the tire is about to slip (such as slipping caused by driving into a water accumulation area or other abnormal road surfaces).

[0094] In the embodiment of the present application, the pre-trained emergency event detection model is trained by a long short-term memory network model. A neural network is trained with a large amount of experimental data (including acoustic signals during normal driving and sudden slipping), and this network can remember the temporal rules of sound features (for example: the high-frequency energy continuously decreases within 300 milliseconds, and at the same time the resonance frequency suddenly jitters), providing support for the detection of emergency events.

[0095] Input the key features of the current and the past 300 milliseconds into the model to output a mutation probability value. When the probability exceeds the fourth threshold (such as 95%), it is considered that the tire is about to lose adhesion (for example: the front wheel of the vehicle has contacted the water accumulation, and the rear wheel is still on the dry road surface. At this time, the torque of the rear wheel needs to be limited in advance).

[0096] Further, step S4 includes:

[0097] Obtain the sound features of different types of road surfaces and their corresponding actual friction coefficients;

[0098] Construct a mapping relationship between the sound features of different types of road surfaces and the corresponding actual friction coefficients;

[0099] When the detection result is encountering an emergency event, input the key features into a pre-constructed road surface adhesion prediction model to obtain the predicted friction coefficient.

[0100] In the above implementation process, the basic situation of the road surface is obtained in advance according to the mapping relationship between the sound features of different road surfaces and their actual friction coefficients, which is convenient for accurately predicting the friction coefficient, reducing the data error generated during the prediction process, and reducing the interference of abnormal road surface conditions.

[0101] In the embodiments of the present application, the vehicle is made to drive over different types of road surfaces such as dry land, wet land, and ice at different speeds through pre-tests, and the corresponding sound characteristics and actual friction coefficients are recorded. The actual friction coefficient can be deduced by measuring the tire slip rate.

[0102] According to Gaussian process regression, a mathematical relationship (mapping relationship) between the sound characteristics and the actual friction coefficient is constructed. For example, a 30% decrease in high-frequency energy may correspond to an actual friction coefficient μ = 0.3 (wet and slippery road surface), and at the same time, the confidence interval of the predicted friction coefficient is given (such as μ = 0.3 ± 0.05).

[0103] When an adhesion mutation is detected, the current key features are input into the model, and the predicted friction coefficient is output.

[0104] Further, S5 includes:

[0105] Obtain the torque limit data corresponding to the predicted friction coefficient;

[0106] When the torque limit data is less than the fifth threshold, perform cooperative pre-boosting on the electric vehicle and control the torque rising speed to achieve pre-tightening control of the electric vehicle.

[0107] In the above implementation process, cooperative pre-boosting is performed on the electric vehicle according to the torque limit data, so that the electric vehicle can pressurize the brake caliper in advance and give a quick response when braking is required, reducing the safety hazards caused by torque limitation.

[0108] The torque limit data corresponding to different friction coefficients (or) the safety torque limits are different. For example:

[0109] μ = 0.2 (ice surface), limit the maximum increase in driving torque to 50 N·m per second to avoid slipping caused by sudden acceleration; μ = 0.4 (wet asphalt), allow the torque to increase by 100 N·m per second; μ = 0.8 (dry road surface), no torque limit.

[0110] The present application presets the fifth threshold. When the torque limit data is less than the fifth threshold, it can be determined that the vehicle has traveled to an abnormal road surface (or a low-adhesion road surface).

[0111] Send instructions to the motor or engine controller through the vehicle CAN bus (in-vehicle communication network) to limit the torque rising speed. For example, when the driver suddenly steps on the accelerator, the system will "smooth" the accelerator signal to make the torque increase slowly instead of bursting instantly.

[0112] Cooperative pre-boosting means that when a low-adhesion road surface is predicted, the braking system is made to build up a certain pressure (such as 5 MPa) in advance, so as to shorten the response time when the anti-lock braking system is triggered. In the embodiments of the present application, an instruction is sent to the electronic stability program to pressurize the brake cylinder in advance (without actual braking) to ensure a quick response when braking is required.

[0113] Further, after the step of performing pre-tightening control on the electric vehicle according to the predicted friction coefficient, the method further includes:

[0114] Obtaining the actual slip ratio of the electric vehicle;

[0115] Performing error correction on the predicted friction coefficient according to the actual slip ratio to obtain an error rate;

[0116] Adjusting the parameters of the road surface adhesion prediction model according to the error rate.

[0117] In the above implementation process, the actual friction data during the driving of the electric vehicle is inversely deduced according to the actual slip ratio of the electric vehicle, and the predicted friction data is corrected for errors according to the actual friction data, and the parameter settings of the road surface adhesion prediction model are adjusted, which can improve the accuracy of the road surface adhesion prediction model and the accuracy of the predicted friction coefficient.

[0118] The actual slip ratio is obtained by measuring the rotational speed difference of the four wheels through a wheel speed sensor. For example, when the rotational speed of the driving wheel is much greater than that of the non-driving wheel, it indicates that the tire is slipping.

[0119] Comparing the predicted friction coefficient with the true friction coefficient deduced from the actual slip ratio. If the error exceeds 15% (error rate), the parameters of the road surface adhesion prediction model are automatically adjusted.

[0120] The embodiments of the present application can also record the operation habits of the driver under different road conditions (such as the throttle opening under a wet and slippery road surface) to assist in optimizing the control strategy.

[0121] Exemplarily, the vehicle enters a locally waterlogged area of the road surface at 80 km / h.

[0122] Control timing:

[0123] 1. t = 0 ms: The microphone detects a 40% decrease in the energy in the 4 kHz frequency band, triggering the detection of adhesion mutation;

[0124] 2. t + 50 ms: The friction coefficient μ predicted by the GPR model is 0.25, entering the low-adhesion control mode;

[0125] 3. t + 100 ms: Limiting the driving torque gradient to 50 Nm / ms and pre-boosting to 3 MPa;

[0126] 4.t + 300 ms: The vehicle enters the water accumulation area, and the actual slip ratio λ = 0.12 (< safety threshold 0.15).

[0127] Through methods such as acoustic timing analysis and prediction, the embodiment of the present application can identify sudden changes in tire adhesion 200 - 300 ms in advance, leaving a key decision-making window for vehicle control. Only 4 directional microphones are required, without high-value devices such as lidar and rain sensors, and the hardware cost is low.

[0128] Based on the torque gradient dynamic limiting strategy of predicted friction coefficient, the embodiment of the present application can suppress the fluctuation of the drive wheel slip ratio (the peak-to-peak value is reduced by 40%), and avoid the sense of power interruption caused by sudden braking.

[0129] Correct the error friction, achieve closed-loop feedback, and adapt to long-term changes such as tire wear and road surface aging.

[0130] Cover complex scenarios, distinguish local sudden low adhesion (such as a single-sided puddle) from continuous slippery road surfaces (icy roads). For the former, only limit the torque of the single-sided wheel, and for the latter, globally reduce the power, solving the triangular contradiction of "perception lag, control conflict, and safety redundancy" during driving, and providing the ability of coupled-state real-time perception and control for low-cost and high-reliability road surfaces.

[0131] Embodiment 2

[0132] In order to execute the method corresponding to Embodiment 1 above to achieve the corresponding functions and technical effects, the following provides a pre-tightening control device for an electric vehicle, as Figure 2 shown. The device includes:

[0133] An acquisition module 1 for acquiring the acoustic signal of the friction between the electric vehicle tire and the ground;

[0134] An extraction module 2 for extracting key features from the preprocessed acoustic signal to obtain key features;

[0135] A detection module 3 for detecting emergencies according to the key features to obtain a detection result;

[0136] A prediction module 4 for predicting the road surface adhesion according to the key features to obtain the predicted friction coefficient if the detection result is that an emergency is encountered;

[0137] A pre-tightening control module 5 for performing pre-tightening control on the electric vehicle according to the predicted friction coefficient.

[0138] In the above implementation process, the detection of emergencies and the prediction of road surface adhesion are carried out in sequence according to the key features. The pre-tightening control of the electric vehicle is performed according to the predicted friction coefficient. The pre-tightening control method can be given in a timely manner according to the captured road surface changes, reducing the interference of abnormal road surfaces on driving, with faster response, more stable and effective control of the electric vehicle, less dependence on hardware, and the ability to meet various driving scenarios.

[0139] Furthermore, the extraction module 2 is also used for:

[0140] Segment the preprocessed acoustic signal to obtain multiple segmented sound signals;

[0141] Analyze the multiple segmented sound signals respectively to obtain the sound intensities of multiple different frequencies;

[0142] Determine the ratio of the sound intensity in the first threshold frequency range to the sound intensities of multiple different frequencies as the high-frequency energy ratio;

[0143] Detect the change in the resonance frequency of the tire in the preprocessed acoustic signal according to the second threshold frequency range to obtain the resonance peak offset rate;

[0144] Extract the transient impact count of the tire from the preprocessed acoustic signal according to the third threshold;

[0145] Determine the high-frequency energy ratio, resonance peak offset rate, and transient impact count as key features.

[0146] In the above implementation process, different key features are extracted from the acoustic signal respectively to obtain the acoustic performance of the acoustic signal under different conditions, which can truly reflect the road surface changes during the driving of the electric vehicle and provide data support for the pre-tightening control.

[0147] Furthermore, the detection module 3 is also used for:

[0148] Input the key features into a pre-trained emergency detection model for detection to obtain a mutation probability value;

[0149] Judge whether the mutation probability value is greater than the fourth threshold;

[0150] If so, determine the detection result as encountering an emergency.

[0151] In the above implementation process, the detection of emergencies for the key features can detect whether there are mutations on the road surface during driving, reduce the impact of abnormal road surfaces on the driving of the electric vehicle, improve the robustness of the vehicle, and provide safety guarantee for the pre-tightening control of the electric vehicle.

[0152] Furthermore, the prediction module 4 is also used for:

[0153] Obtain the sound characteristics of different types of road surfaces and their corresponding actual friction coefficients;

[0154] Construct a mapping relationship between the sound characteristics of different types of road surfaces and their corresponding actual friction coefficients;

[0155] When the detection result is encountering an emergency, input the key features into the pre-constructed road surface adhesion prediction model to obtain the predicted friction coefficient.

[0156] In the above implementation process, the basic situation of the road surface is obtained in advance according to the mapping relationship between the sound characteristics of different road surfaces and their actual friction coefficients, which is convenient for accurately predicting the friction coefficient, reducing the data error generated in the prediction process, and reducing the interference of abnormal road surface conditions.

[0157] Furthermore, the pre-tightening control module 5 is also used for:

[0158] Obtain the torque limit data corresponding to the predicted friction coefficient;

[0159] When the torque limit data is less than the fifth threshold, perform coordinated pre-boosting on the electric vehicle and control the torque rising speed to achieve pre-tightening control of the electric vehicle.

[0160] In the above implementation process, coordinated pre-boosting is performed on the electric vehicle according to the torque limit data, so that the electric vehicle can pressurize the brake caliper in advance and give a quick response when braking is required, reducing the safety hazard caused by torque limitation.

[0161] Furthermore, the device also includes a feedback module for:

[0162] Obtain the actual slip ratio of the electric vehicle;

[0163] Perform error correction on the predicted friction coefficient according to the actual slip ratio to obtain the error rate;

[0164] Adjust the parameters of the road surface adhesion prediction model according to the error rate.

[0165] In the above implementation process, the actual friction data during the driving of the electric vehicle is deduced based on the actual slip ratio of the electric vehicle, and the predicted friction data is corrected for errors according to the actual friction data, and the parameter settings of the road surface adhesion prediction model are adjusted, which can improve the accuracy of the road surface adhesion prediction model and the accuracy of the predicted friction coefficient.

[0166] The above pre-tightening control device for electric vehicles can implement the method of the first embodiment above. The optional items in the first embodiment above also apply to this embodiment and will not be elaborated here.

[0167] The remaining content of the embodiments of the present application can refer to the content of the first embodiment above and will not be repeated in this embodiment.

[0168] Embodiment III

[0169] An embodiment of the present application provides an electric vehicle, including the pre-tightening control device of the electric vehicle in Embodiment II.

[0170] Embodiment IV

[0171] An embodiment of the present application provides an electronic device, including a memory and a processor. The memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the pre-tightening control method of the electric vehicle in Embodiment I.

[0172] Optionally, the above-mentioned electronic device may be a server.

[0173] Please refer to Figure 3 , Figure 3 , which is a schematic diagram of the structural composition of the electronic device provided by the embodiment of the present application. The electronic device may include a processor 31, a communication interface 32, a memory 33, and at least one communication bus 34. Among them, the communication bus 34 is used to realize the direct connection communication between these components.

[0174] Optionally, the electronic device may further include a storage controller and an input / output unit. Each component of the memory 33, the storage controller, the processor 31, the peripheral interface, and the input / output unit is directly or indirectly electrically connected to each other to realize data transmission or interaction.

[0175] The input / output unit is used to provide the user to create a task and create a start optional period or a preset execution time for the task to realize the interaction between the user and the server. The input / output unit may be, but is not limited to, a mouse, a keyboard, etc.

[0176] It can be understood that Figure 3 The structure shown is only schematic, and the electronic device may further include more or fewer components than those shown in Figure 3 , or have a different configuration from that shown in Figure 3 . Figure 3 Each component shown in

[0177] can be implemented by hardware, in-vehicle software, or a combination thereof.

[0178] In addition, an embodiment of the present application further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, it realizes the pre-tightening control method of the electric vehicle in Embodiment I.

[0179] The above are only the embodiments of the present application and are not used to limit the protection scope of the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application. It should be noted that similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0180] As described above, the above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, and all should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A preload control method for an electric vehicle, characterized in that: The method comprises: Acquire the acoustic signal of the friction between the tires of the electric vehicle and the ground; Extract key features from the preprocessed acoustic signal to obtain key features; Perform emergency event detection according to the key features to obtain detection results; If the detection result is an unexpected event, a road adhesion prediction is performed based on the key features to obtain a predicted friction coefficient; The electric vehicle is preloaded and controlled according to the predicted friction coefficient.

2. The preload control method of an electric vehicle according to claim 1, characterized in that: The step of extracting key features from the preprocessed acoustic signal to obtain key features includes: Segmenting the preprocessed acoustic signal to obtain a plurality of segmented sound signals; Analyzing the multiple segmented sound signals respectively to obtain multiple sound intensities of different frequencies; Determine the ratio of the sound intensity of the first threshold frequency range in the sound intensities of the plurality of different frequencies as the high frequency energy ratio; Detecting a change in the resonance frequency of the tire in the preprocessed acoustic signal according to a second threshold frequency range to obtain a resonance peak shift rate; extracting a transient impact count of the tire from the preprocessed acoustic signal according to a third threshold; The high frequency energy ratio, the resonance peak shift rate and the transient impulse count are determined as the key features.

3. The preload control method of an electric vehicle according to claim 1, characterized in that: The step of performing emergency event detection according to the key features to obtain the detection result comprises: Inputting the key features into a pre-trained emergency detection model for detection to obtain a mutation probability value; Determining whether the mutation probability value is greater than a fourth threshold; If so, it is determined that the detection result is an emergency event.

4. The preload control method of an electric vehicle according to claim 1, characterized in that: The step of predicting road adhesion according to the key features to obtain a predicted friction coefficient comprises: Obtain the sound characteristics of different types of road surfaces and their corresponding actual friction coefficients; Construct a mapping relationship between the sound characteristics of different types of road surfaces and the corresponding actual friction coefficients; When the detection result is that an emergency event is encountered, the key feature is input into a pre-built road adhesion prediction model to obtain the predicted friction coefficient.

5. The preload control method of an electric vehicle according to claim 1, characterized in that: The step of performing preload control on the electric vehicle according to the predicted friction coefficient comprises: Obtaining torque limit data corresponding to the predicted friction coefficient; When the torque limit data is less than a fifth threshold, the electric vehicle is pre-pressurized in a coordinated manner, and a torque increase speed is controlled to achieve pre-tightening control of the electric vehicle.

6. The preload control method of an electric vehicle according to claim 5, characterized in that: After the step of performing preload control on the electric vehicle according to the predicted friction coefficient, the method further includes: Obtaining an actual slip rate of the electric vehicle; Performing error correction on the predicted friction coefficient according to the actual slip rate to obtain an error rate; The parameters of the road adhesion prediction model are adjusted according to the error rate.

7. A preload control device for an electric vehicle, characterized in that: The device comprises: An acquisition module, used to acquire acoustic signals of friction between the tire of the electric vehicle and the ground; An extraction module is used to extract key features from the preprocessed acoustic signal to obtain key features; A detection module, used to detect emergencies according to the key features and obtain detection results; A prediction module, for predicting road adhesion according to the key features to obtain a predicted friction coefficient if the detection result is an unexpected event; A preload control module is used to perform preload control on the electric vehicle according to the predicted friction coefficient.

8. An electric vehicle, characterized in that: The electric vehicle comprises the preload control device for the electric vehicle according to claim 7.

9. An electronic device, characterized in that: The electronic device comprises a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the preload control method of the electric vehicle according to any one of claims 1 to 6.

10. A storage medium, characterized in that: It stores a computer program, and when the computer program is executed by a processor, the preload control method of the electric vehicle as claimed in any one of claims 1 to 6 is implemented.

Images