A Shielding and Noise Reduction Method and System for the High-Voltage Harness of a New Energy Vehicle

Abstract

The present invention relates to the field of electromagnetic compatibility technology, and discloses a method and system for shielding and noise reduction of high-voltage wiring harnesses of new energy vehicles. The method includes obtaining current fluctuation data, working state codes, and environmental condition data; performing wavelet decomposition on the current fluctuation data to obtain current fluctuation characteristics; performing environmental analysis on the environmental condition data to obtain environmental state characteristics; inputting the current fluctuation characteristics, the working state codes, and the environmental state characteristics into a preset current prediction model to obtain predicted current data; extracting noise according to the predicted current data to obtain high-frequency noise data; and performing impedance matching according to the high-frequency noise data to output target impedance parameters. The present method has the following effects: The present method can improve the shielding and noise reduction accuracy of high-voltage wiring harnesses of new energy vehicles.

Description

Technical Field

[0001] The present invention relates to the technical field of electromagnetic compatibility, and particularly to a method and system for shielding and noise reduction of high-voltage harnesses in new energy vehicles. Background Art

[0002] At present, as an alternative to traditional fuel vehicles, new energy vehicles are gradually becoming the mainstream means of transportation. The high-voltage harness is one of the crucial components in new energy vehicles, responsible for transmitting power and signals between various electrical systems of the vehicle. Due to its high operating voltage and large current, the design of the high-voltage harness not only needs to consider electrical performance but also pay special attention to electromagnetic compatibility (EMC) and noise reduction effects. Especially in the context of the increasing number of electronic devices in modern vehicles, it is particularly important to ensure that the high-voltage harness can effectively resist external electromagnetic interference and reduce the noise generated by itself. In addition, the high-voltage harness must also have good mechanical strength and durability to meet the requirements of complex driving environments and long-term use. Therefore, the research and development of high-performance shielding and noise reduction technologies are of great significance for improving the overall performance of new energy vehicles.

[0003] In an existing technology, a shielding and noise reduction method is to add one or more layers of metal shielding layers outside the high-voltage harness, and on this basis, combine an absorbing layer made of magnetic materials to jointly construct a composite protection structure. The specific steps are as follows: First, select a suitable metal material, such as aluminum foil or copper braid, and wrap it around the high-voltage cable to form the first line of defense, mainly used to block high-frequency electromagnetic interference signals. Then, apply an absorbing layer composed of ferrite or other magnetic materials outside the metal shielding layer. This step aims to further weaken the intensity of electromagnetic waves in the low-frequency band and prevent them from penetrating the shielding layer to cause interference. In addition, to enhance the reliability of the overall system, filters are also installed at both ends of the cable to eliminate common-mode noise on the conduction path.

[0004] Although the above shielding and noise reduction methods can provide good protection in theory, their shortcoming is the lack of dynamic matching ability with environmental conditions and the operating state of the vehicle body. For example, under different driving conditions (such as acceleration, deceleration, turning, etc.), the load of the in-vehicle electrical system will change significantly, resulting in changes in the frequency and intensity of the generated electromagnetic interference. However, the existing shielding solutions cannot adjust their protection strategies in real time according to these changes, resulting in poor EMI suppression effects under different working conditions. At the same time, the external environment during vehicle driving also has an important impact on electromagnetic interference. For example, when approaching a substation or a radio transmitting tower, the external strong electromagnetic field will exceed the original design range, making the originally effective shielding measures no longer applicable. More importantly, the complex electromagnetic environment changes caused by the interaction between various subsystems inside the vehicle are not considered, which also limits the applicability of the existing technology in a wider range of scenarios. In summary, the accuracy of the existing high-voltage wire harness shielding and noise reduction technology is low. Summary of the Invention

[0005] The present invention provides a shielding and noise reduction method and system for a high-voltage wire harness of a new energy vehicle to improve the shielding and noise reduction accuracy of the high-voltage wire harness.

[0006] In a first aspect, to solve the above technical problems, the present invention provides a shielding and noise reduction method for a high-voltage wire harness of a new energy vehicle, including:

[0007] Obtain current fluctuation data, working state codes, and environmental condition data;

[0008] Perform wavelet decomposition on the current fluctuation data to obtain current fluctuation characteristics;

[0009] Perform environmental analysis on the environmental condition data to obtain environmental state characteristics;

[0010] Input the current fluctuation characteristics, the working state code, and the environmental state characteristics into a preset current prediction model to obtain predicted current data;

[0011] Extract noise based on the predicted current data to obtain high-frequency noise data;

[0012] Perform impedance matching based on the high-frequency noise data and output target impedance parameters.

[0013] In an optional implementation manner, the obtaining of the current fluctuation data, the working state code, and the environmental condition data includes:

[0014] Obtain the electromagnetic interference frequency;

[0015] When the electromagnetic interference frequency is less than a preset first sampling frequency, perform data acquisition according to the first sampling frequency to obtain current fluctuation data;

[0016] When the electromagnetic interference frequency is greater than or equal to the first sampling frequency, frequency update is performed according to the electromagnetic interference frequency to obtain a second sampling frequency;

[0017] When the electromagnetic interference frequency is greater than the second sampling frequency, data acquisition is performed according to the second sampling frequency to obtain current fluctuation data;

[0018] Obtain the vehicle acceleration;

[0019] When the vehicle acceleration is greater than a preset first working threshold, determine that the working state code is a preset first code;

[0020] When the vehicle acceleration is less than or equal to the first working threshold and greater than a preset second working threshold, determine that the working state code is a preset second code;

[0021] When the vehicle acceleration is less than or equal to the second working threshold, determine that the working state code is a preset third code.

[0022] In an alternative embodiment, the performing wavelet decomposition on the current fluctuation data to obtain current fluctuation features includes:

[0023] Perform basis function matching on the current fluctuation data to obtain a decomposition basis function;

[0024] Perform recursive decomposition on the decomposition basis function and the current fluctuation data to obtain a set of subband coefficients;

[0025] Perform signal reconstruction on the set of subband coefficients to obtain subband time-domain signals;

[0026] Perform entropy quantization on the subband time-domain signals to obtain current fluctuation features.

[0027] In an alternative embodiment, the performing environmental analysis on the environmental condition data to obtain environmental state features includes:

[0028] Perform spectrum analysis on the environmental condition data to obtain the high-frequency energy ratio;

[0029] When the high-frequency energy ratio is greater than a preset high-frequency threshold, determine that the electromagnetic propagation path is radiation propagation;

[0030] Perform periodic detection on the environmental condition data to obtain a propagation period;

[0031] When the propagation period is an integer, determine that the electromagnetic propagation path is waveguide propagation;

[0032] Extract the field strength according to the electromagnetic propagation path and the environmental condition data to obtain the field strength data;

[0033] Perform feature fusion according to the electromagnetic propagation path and the field strength data to obtain the environmental state features.

[0034] In an alternative embodiment, the training process of the current prediction model includes:

[0035] Construct a current prediction model based on a long short-term memory neural network, where the input layer includes historical current features, historical working states, and historical environmental features;

[0036] The output layer is the target current prediction value;

[0037] Use the mean square error between the predicted value and the true value as the loss function during the iteration process;

[0038] Complete one round of iteration and update the parameters;

[0039] When it is detected that the loss function of the model meets the conditions or the number of training times reaches the set upper limit, determine that the training is completed and obtain the trained model.

[0040] In an alternative embodiment, the extraction of high-frequency noise data according to the predicted current data includes:

[0041] Perform baseline correction on the predicted current data to obtain residual waveform data;

[0042] Perform frequency domain transformation on the residual waveform data to obtain the energy spectral density distribution;

[0043] Perform high-frequency filtering on the energy spectral density distribution to obtain the high-frequency noise distribution;

[0044] Perform time-frequency joint annotation on the high-frequency noise distribution to obtain high-frequency noise data.

[0045] In an alternative embodiment, the impedance matching according to the high-frequency noise data to output the target impedance parameter includes:

[0046] Perform spectrum extraction on the high-frequency noise data to obtain a noise spectrum map;

[0047] Calculate the sensitivity according to the noise spectrum map to obtain the sensitivity distribution curve;

[0048] Perform deviation calculation according to the sensitivity distribution curve to obtain the impedance deviation index;

[0049] Perform multi-parameter optimization according to the impedance deviation index to obtain the target impedance parameter.

[0050] In a second aspect, the present invention provides a shielding and noise reduction system for a high-voltage harness of a new energy vehicle, comprising:

[0051] A data acquisition module, configured to acquire current fluctuation data, working state encoding, and environmental condition data;

[0052] A current feature module, configured to perform wavelet decomposition on the current fluctuation data to obtain current fluctuation features;

[0053] An environmental feature module, configured to perform environmental analysis on the environmental condition data to obtain environmental state features;

[0054] A current prediction module, configured to input the current fluctuation features, the working state encoding, and the environmental state features into a preset current prediction model to obtain predicted current data;

[0055] A noise extraction module, configured to extract noise based on the predicted current data to obtain high-frequency noise data;

[0056] An impedance matching module, configured to perform impedance matching based on the high-frequency noise data and output target impedance parameters.

[0057] In a third aspect, the present invention further provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the shielding and noise reduction method for the high-voltage harness of a new energy vehicle described in any one of the above is implemented.

[0058] In a fourth aspect, the present invention further provides a computer-readable storage medium, which includes a stored computer program. When the computer program runs, it controls the device where the computer-readable storage medium is located to execute the shielding and noise reduction method for the high-voltage harness of a new energy vehicle described in any one of the above.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] (1) The process of acquiring current fluctuation data, working state encoding, and environmental condition data ensures comprehensive monitoring of the system operation status. By accurately collecting these data, a solid foundation is provided for subsequent analysis. This step not only improves the integrity and accuracy of the data but also provides high-quality data input for subsequent wavelet decomposition, environmental analysis, etc., thereby enhancing the reliability and stability of the overall system.

[0061] (2)Perform wavelet decomposition on the said current fluctuation data to obtain current fluctuation characteristics. The wavelet decomposition method can effectively extract key feature information from complex current fluctuation signals, including the energy distribution in different frequency bands, etc. This step significantly improves the ability to understand the current fluctuation pattern, helps to more accurately identify potential problems, provides accurate input data for the current prediction model, and enhances the reliability of the prediction results.

[0062] (3)Perform environmental analysis on the said environmental condition data to obtain environmental state characteristics. By carefully analyzing the environmental conditions, various external factors affecting the system performance can be captured. This analysis helps to adjust the system parameters to adapt to different working environments and reduce the impact of environmental changes on the system performance. As one of the important input information, the environmental state characteristics play a key role in improving the accuracy of the current prediction model.

[0063] (4)Input the said current fluctuation characteristics, the working state encoding, and the environmental state characteristics into a preset current prediction model to obtain predicted current data. By integrating multiple feature information into the current prediction model, the system can more accurately simulate the actual current change trend. This method not only improves the prediction accuracy but also better reflects the response characteristics of the system under different working conditions, providing a reliable basis for subsequent noise extraction.

[0064] (5)Perform noise extraction on the said predicted current data to obtain high-frequency noise data. Use advanced signal processing techniques to separate the high-frequency noise components from the predicted current data. This process helps to identify the interference sources in the system and their influence levels. Accurately extracting high-frequency noise data is crucial for subsequent impedance matching because it directly relates to whether the noise can be effectively suppressed to ensure the stable operation of the system.

[0065] (6)Perform impedance matching based on the said high-frequency noise data and output the target impedance parameters. By analyzing the high-frequency noise data and performing impedance matching, the electrical characteristics of the system can be optimized, and the impact of noise on the system performance can be reduced. This method can not only enhance the anti-interference ability of the system but also improve the quality of the current waveform and the overall working efficiency. Finally, based on the calculated target impedance parameters for adjustment, the optimal operating state of the system can be achieved to meet the requirements of high-performance applications. Description of the Drawings

[0066] Figure 1 is a schematic flowchart of a shielding and noise reduction method for a high-voltage harness of a new energy vehicle provided by the first embodiment of the present invention;

[0067] Figure 2 is a schematic structural diagram of a shielding and noise reduction system for a high-voltage harness of a new energy vehicle provided by the second embodiment of the present invention. Specific implementation manner

[0068] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0069] Refer to Figure 1 , a shielding and noise reduction method for a high-voltage wire harness of a new energy vehicle according to a first embodiment of the present invention includes the following steps:

[0070] S11, obtaining current fluctuation data, a working state code, and environmental condition data;

[0071] S12, performing wavelet decomposition on the current fluctuation data to obtain current fluctuation characteristics;

[0072] S13, performing environmental analysis on the environmental condition data to obtain environmental state characteristics;

[0073] S14, inputting the current fluctuation characteristics, the working state code, and the environmental state characteristics into a preset current prediction model to obtain predicted current data;

[0074] S15, extracting noise according to the predicted current data to obtain high-frequency noise data;

[0075] S16, performing impedance matching according to the high-frequency noise data and outputting target impedance parameters.

[0076] In step S11, current fluctuation data, a working state code, and environmental condition data are obtained.

[0077] It should be noted that the current fluctuation data is collected in real time by the Hall sensor array terminated at the high-voltage harness, recording the transient current changes during motor operation at a sampling frequency of 1 MHz and stored as a CSV time-series data file with timestamps (such as 20230313_1530_BatteryCurrent.csv, containing millisecond-level timestamps and current values in the range of ±500 A); the working state code is transmitted by the vehicle controller via the CAN bus, using the PGN63488 message in the J1939 protocol, and representing the vehicle operation mode in hexadecimal encoding (e.g., 0x01 represents the acceleration condition, 0x02 represents the energy recovery condition), stored as a binary log file with the message ID; the environmental condition data is obtained through the temperature and humidity sensors and EMI detection module integrated in the BMS system, including the temperature gradient from -40°C to 125°C, humidity values from 0 to 100% RH, and electromagnetic field strength in the 0 - 200 MHz frequency band, stored as a structured record table in the SQLite database, and each record contains 20 fields such as longitude and latitude coordinates, timestamps, and three-axis magnetic field strength. The three types of data are time-synchronized through the in-vehicle gateway to form a fusion acquisition system for multi-source heterogeneous data.

[0078] In one implementation, the electromagnetic interference frequency is obtained; when the electromagnetic interference frequency is less than a preset first sampling frequency, data acquisition is performed according to the first sampling frequency to obtain current fluctuation data; when the electromagnetic interference frequency is greater than or equal to the first sampling frequency, frequency update is performed according to the electromagnetic interference frequency to obtain a second sampling frequency; when the electromagnetic interference frequency is greater than or equal to the second sampling frequency, data acquisition is performed according to the second sampling frequency to obtain current fluctuation data; the vehicle acceleration is obtained; when the vehicle acceleration is greater than a preset first working threshold, the working state code is determined as a preset first code; when the vehicle acceleration is less than or equal to the first working threshold and greater than a preset second working threshold, the working state code is determined as a preset second code; when the vehicle acceleration is less than or equal to the second working threshold, the working state code is determined as a preset third code.

[0079] It should be noted that the key parameters involved in this embodiment are analyzed as follows: The electromagnetic interference frequency refers to the main frequency band of electromagnetic noise generated during the operation of the high-voltage system. For example, the 20 kHz fundamental frequency generated by the PWM modulation of the drive motor and its third harmonic 60 kHz; The first sampling frequency is the preset basic data acquisition rate (such as 50 kHz). When the detected interference frequency reaches 80 kHz, the second sampling frequency (such as 100 kHz) is enabled for anti-aliasing sampling. The vehicle acceleration is obtained through a vehicle longitudinal acceleration sensor with a range of ±1g. When the acceleration exceeds 0.3g (the first working threshold), it is marked as the rapid acceleration state (the first code 0x01). When it is between 0.1g (the second working threshold) and 0.3g, it is in the normal working condition (the second code 0x02). When it is lower than 0.1g, it is determined as the coasting state (the third code 0x03). This parameter system realizes a dynamic sampling strategy. For example, when the 100 kHz motor harmonic is detected, the sampling rate is automatically increased to 200 kHz. At the same time, the high-frequency data acquisition mode is triggered during the rapid acceleration stage to ensure the capture of the current ripple caused by the IGBT switching transient.

[0080] It should be noted that in the subsequent current prediction model, the first, second, and third codes, as the core characterization parameters of the vehicle working condition, affect the prediction accuracy through a multi-dimensional mechanism: At the physical mapping level of the code, the first code (0x01) corresponds to the high-frequency current harmonic offset caused by the motor IGBT switching frequency rising above 15 kHz under the rapid acceleration working condition. The second code (0x02) represents the stable current ripple range of ±50A corresponding to a 40 - 70% load rate during normal driving. The third code (0x03) reflects the current direction reversal and the 60 kHz characteristic interference in the energy recovery coasting mode; At the feature coupling level, a 32-dimensional LSTM input matrix is jointly constructed by the three-dimensional vector transformed by one-hot encoding, the current features of 8-layer wavelet decomposition, and the normalized environmental parameters. Among them, the weight ratio of the working condition code reaches 18%; In the time series modeling, the change of the code sequence within a 500 ms sliding window triggers the state transition detection. For example, three consecutive 0x01 codes will enhance the prediction sensitivity of the model to transient overcurrent.

[0081] In step S12, wavelet decomposition is performed on the current fluctuation data to obtain current fluctuation features.

[0082] In one embodiment, basis function matching is performed on the current fluctuation data to obtain a decomposition basis function; recursive decomposition is performed according to the decomposition basis function and the current fluctuation data to obtain a set of sub-band coefficients; signal reconstruction is performed according to the set of sub-band coefficients to obtain a sub-band time-domain signal; entropy quantization is performed on the sub-band time-domain signal to obtain current fluctuation features.

[0083] It should be noted that in the analysis of current fluctuations, it is first necessary to select a basis function that matches the signal characteristics as the decomposition tool. The selection of the basis function is based on the morphological characteristics of the current data. For example, if the current waveform exhibits short-time pulse characteristics (such as the transient caused by IGBT switching), a basis function with high local time-domain resolution (such as the Daubechies series) is adopted; if the signal has periodic oscillations (such as PWM modulation harmonics), a basis function with good symmetry (such as the Symlets series) is preferably selected. After the matching is completed, the system locks the optimal decomposition basis function, which is essentially a set of waveform templates with specific time-frequency characteristics for subsequent multi-scale signal decomposition.

[0084] It should be noted that based on the selected basis function, the system performs multi-level recursive decomposition on the current fluctuation data. Each level of decomposition disassembles the signal into a low-frequency approximation part (reflecting trend changes) and a high-frequency detail part (capturing transient disturbances). For example, when performing 8-level decomposition, the first level corresponds to the highest-frequency noise (such as switching noise), and the eighth level corresponds to the lowest-frequency load fluctuation trend. The set of subband coefficients is composed of the high- and low-frequency coefficients obtained from each level of decomposition, and its structure shows a tree-like distribution. Each coefficient characterizes the energy distribution and temporal correlation characteristics of the signal in a specific frequency band. For example, a sudden increase in the amplitude of the high-frequency coefficient indicates a motor stall event.

[0085] It should be noted that through the inverse transform algorithm, the system restores the set of subband coefficients to a time-domain signal that can be intuitively analyzed. Each subband signal corresponds to the component of the original current fluctuation in a specific frequency band. For example, the reconstructed third-level subband signal contains PWM carrier harmonics in the range of 10 - 20 kHz, while the sixth-level subband signal reflects the low-frequency oscillation in the order of hundreds of hertz caused by motor torque ripple. These subband time-domain signals not only retain the original temporal characteristics but also enhance the identifiability of specific interference components through frequency band isolation. For example, the periodic current distortion caused by the inverter dead-time effect can be clearly separated.

[0086] It should be noted that finally, the system calculates the entropy value of the subband time-domain signal to quantify its complexity and randomness. For example, calculating the energy entropy for the high-frequency subband (such as the second level) can characterize the burst intensity of the switching noise; calculating the approximate entropy for the low-frequency subband (such as the seventh level) is used to evaluate the smoothness of the load fluctuation. These entropy values together form the current fluctuation feature vector. For example, the energy entropy of the high-frequency subband increases significantly under rapid acceleration conditions, while the approximate entropy of the low-frequency subband tends to be stable during coasting. This feature vector is finally input into the prediction model to achieve accurate prediction of abnormal current fluctuations (such as resonance overshoot) under different working conditions.

[0087] In step S13, environmental analysis is performed based on the environmental condition data to obtain environmental state characteristics.

[0088] In one implementation, spectrum analysis is performed based on the environmental condition data to obtain the high-frequency energy ratio. When the high-frequency energy ratio is greater than a preset high-frequency threshold, it is determined that the electromagnetic propagation path is radiation propagation. Periodic detection is performed based on the environmental condition data to obtain the propagation period. When the propagation period is an integer, it is determined that the electromagnetic propagation path is waveguide propagation. Field strength extraction is performed based on the electromagnetic propagation path and the environmental condition data to obtain field strength data. Feature fusion is performed based on the electromagnetic propagation path and the field strength data to obtain environmental state features.

[0089] It should be noted that the system first scans the spectrum energy distribution of environmental condition data (such as electromagnetic interference spectrum, temperature and humidity, etc.), and focuses on calculating the proportion of energy in the high-frequency band (for example, above 1 GHz) in the total energy. The high-frequency energy ratio reflects the activity level of radiation-type interference in the environment. For example, when the radar module in a vehicle scenario is working, this value will be significantly increased. When this ratio exceeds the preset threshold (set to 30%), the system determines that the current electromagnetic interference is mainly propagated through radiation, manifested as spatial electromagnetic field coupling. Typical scenarios include spatial radiation generated by the unshielded section of high-voltage wiring harnesses.

[0090] It should be noted that for the electromagnetic fluctuation component in the environmental data, the system uses the autocorrelation algorithm to detect its periodicity. If the detected propagation period shows integer characteristics (for example, the period is exactly equal to the reciprocal of the resonance frequency of the vehicle's metal cavity), it is determined to be the waveguide propagation mode. Such propagation often occurs in waveguides formed by structures such as body gaps and wiring harness channels. Its characteristic is that electromagnetic energy propagates directionally along the conductor surface or in a closed space. For example, the crosstalk between the motor controller and the battery pack often shows fluctuations that are integer multiples of a period of 2 ms.

[0091] It should be noted that according to the identified propagation path, the system differentially extracts field strength data: for radiation propagation, it focuses on the peak value of the spatial field strength and its distribution gradient (such as the field strength in the dashboard area reaches 120 V / m); for waveguide propagation, it extracts the field strength attenuation curve on the conductor surface (such as a 15 dB attenuation per meter along the shielded wiring harness). At the same time, the field strength value is corrected by combining environmental temperature and humidity data. For example, when high temperature causes a change in the dielectric constant of the insulating material, the system automatically compensates for the field strength measurement error.

[0092] It should be noted that finally, the system fuses the propagation path identifier (radiation / waveguide) with the corrected field strength data to generate an environmental state feature vector. For example: in a scenario dominated by radiation propagation, the feature vector includes the high-frequency energy ratio, spatial field strength gradient, and temperature and humidity coupling coefficient; while in a waveguide propagation scenario, parameters such as the periodic stability index, conductor attenuation rate, and resonance frequency deviation are fused. This feature vector can accurately characterize the electromagnetic environment state under complex working conditions (such as increased waveguide leakage due to increased moisture in rainy days), providing an anti-interference correction benchmark for subsequent prediction models.

[0093] In step S14, the current fluctuation feature, the working state encoding, and the environmental state feature are input into a preset current prediction model to obtain predicted current data.

[0094] In one implementation, the training process of the current prediction model includes: constructing a current prediction model based on a long short-term memory neural network, where the input layer includes historical current features, historical working states, and historical environmental features;

[0095] The output layer is the target current prediction value;

[0096] In the iterative process, the mean square error between the predicted value and the true value is used as the loss function;

[0097] Complete one round of iteration and update the parameters;

[0098] When it is detected that the loss function of the model meets the conditions or the number of training times reaches the set upper limit, it is determined that the training is completed, and the trained model is obtained.

[0099] It should be noted that the current prediction model is constructed based on the LSTM model. The dataset of this LSTM current prediction model consists of three parts: current fluctuation features (including sub-band entropy values such as high-frequency energy entropy and low-frequency approximate entropy and time-domain statistics), working state encoding (using one-hot encoding or numerical mapping to represent the device operation mode and control instructions), and environmental state features (including standardized parameters such as temperature and humidity, electromagnetic field strength, and propagation mode identifiers). It is organized in the form of a sliding window (for example, predicting the current value in the next 5 seconds based on the feature sequence in the past 60 seconds), and is trained through a neural network architecture containing 2 - 3 layers of LSTM units (64 - 256 neurons in each layer, with a Dropout rate of 0.2 - 0.5). The Adam optimizer (initial learning rate of 0.001 combined with the cosine annealing strategy) and the composite loss function of MAE and RMSE (weight ratio 6:4) are used, and efficient training is achieved by combining the early stopping method (patience value of 10 epochs) and the model checkpoint mechanism; its purpose is to achieve device protection (microsecond-level abnormal response), energy efficiency optimization (adapting to working conditions to improve efficiency by 5 - 8%), fault warning (identifying device aging 3 - 6 months in advance), and environmental adaptability (automatically compensating for temperature drift and electromagnetic interference) through multi-factor coupling prediction.

[0100] In step S15, high-frequency noise data is obtained by extracting noise from the predicted current data.

[0101] In one implementation, baseline correction is performed on the predicted current data to obtain residual waveform data; frequency domain transformation is performed on the residual waveform data to obtain the energy spectral density distribution; high-frequency filtering is performed on the energy spectral density distribution to obtain the high-frequency noise distribution; time-frequency joint annotation is performed on the high-frequency noise distribution to obtain high-frequency noise data.

[0102] It should be noted that the system analyzes and predicts the long-term change trend of the current (such as minute-level fluctuations) based on a sliding time window, and uses an adaptive weight algorithm to fit a dynamic baseline. By performing baseline correction on the predicted current data, the system first eliminates the steady-state current component under normal operating conditions of the device. The residual waveform data characterizes the instantaneous deviation between the actual current and the predicted baseline, and the abnormal forms such as prominent spikes and oscillations in its waveform can reflect the transient disturbances caused by electromagnetic interference (such as current glitches caused by the switching of power devices). The time resolution of this data is accurate to the microsecond level, and it can clearly capture the aperiodic fluctuation characteristics caused by events such as relay bounce and capacitor charging and discharging, providing the original signal basis for subsequent noise separation.

[0103] It should be noted that by subtracting the original current data from the dynamic baseline point by point, a residual waveform containing only transient fluctuations is obtained. The residual waveform data is converted into an energy spectral density distribution through frequency domain transformation. This distribution uses frequency as the horizontal axis and energy intensity as the vertical axis, intuitively showing the aggregation characteristics of electromagnetic energy in different frequency bands. The low-frequency band (such as 0 - 100 kHz) in the energy spectrum corresponds to the power frequency harmonics of normal device operation, while the energy spikes in the high-frequency band (above 1 MHz) mostly originate from the radiation noise of the switching power supply or the out-of-band leakage of the wireless communication module. By analyzing the steepness and peak position of the spectral density, the frequency characteristics and propagation path of the noise source can be preliminarily judged.

[0104] It should be noted that based on the energy spectral density distribution, the system uses a high-pass filter with an adjustable cut-off frequency to separate the high-frequency noise distribution. This distribution focuses on the 1 MHz - 1 GHz frequency band, and its morphological characteristics include three typical modes: narrowband spikes (such as CAN bus clock harmonics); broadband noise base (such as electromagnetic radiation caused by IGBT switching); periodic comb spectrum (such as sideband leakage generated by PWM modulation). Each distribution mode corresponds to a specific interference mechanism. For example, the energy diffusion degree of the broadband noise base can directly reflect the impact of the switching speed of power devices on electromagnetic compatibility.

[0105] It should be noted that by performing time-frequency joint annotation on the high-frequency noise distribution, structured high-frequency noise data is finally generated. This data contains three core dimensions: time marker (the moment of noise burst accurate to the nanosecond level); frequency marker (the main frequency of the noise and the -3 dB bandwidth); intensity marker (the equivalent radiation intensity in dBμA). In particular, the transient high-frequency noise recorded in the data (such as pulse groups with a duration < 1 μs) can be associated with the parasitic parameter resonance phenomenon of specific circuit nodes, while the persistent high-frequency noise (such as broadband radiation > 10 ms) is mostly related to structural defects such as poor grounding of the radiator. These data provide a quantitative basis for electromagnetic compatibility rectification. For example, by identifying the dense noise clusters in the 200 - 400 MHz frequency band, the radiation over-standard problem caused by the motor drive wire harness without magnetic rings can be located.

[0106] In step S16, impedance matching is performed according to the high-frequency noise data, and target impedance parameters are output.

[0107] In one implementation, spectrum extraction is performed according to the high-frequency noise data to obtain a noise map; sensitivity calculation is performed according to the noise map to obtain a sensitivity distribution curve; deviation calculation is performed according to the sensitivity distribution curve to obtain an impedance deviation index; and multi-parameter optimization is performed according to the impedance deviation index to obtain target impedance parameters.

[0108] It should be noted that based on the high-frequency noise data, spectrum extraction is performed, and the system generates a noise map, which completely presents the noise characteristics in the three-dimensional form of frequency-time-intensity. Different color blocks in the map represent the noise intensity distribution of specific frequency bands. For example, the red area represents continuous radiation noise in the 1-5 MHz band, and the blue stripes correspond to transient pulse groups above 10 MHz. The noise map can intuitively identify the main frequency position of the interference source (such as the fundamental frequency and harmonics of the switching power supply) and its evolution pattern over time (such as the periodic fluctuations caused by PWM modulation), providing a frequency-domain characteristic reference for subsequent impedance matching.

[0109] It should be noted that by analyzing the correlation between the energy ratio of different frequency bands in the noise map and the device function, the system calculates and generates a sensitivity distribution curve. The vertical axis of this curve represents the sensitivity of the device performance to the noise of a specific frequency. For example, the sensitivity peak in the wireless communication band (such as 2.4 GHz) indicates that the noise here will directly interfere with the signal transmission quality. The curve shape can reveal the key vulnerable frequency bands: a steep rising edge indicates that narrowband interference is likely to cause a sharp increase in the bit error rate, and the flat area represents the cumulative effect of broadband noise on the system. The sensitivity data provides a quantitative basis for the impedance optimization priority.

[0110] In one implementation, the sensitivity calculation quantifies the sensitivity of the device performance to electromagnetic noise of different frequencies, and based on the correlation between the energy distribution of each frequency band in the noise map and the core functions of the device (such as wireless communication, signal sampling), a frequency-sensitivity mapping curve is generated, where the peak corresponds to the key vulnerable frequency band (such as the sensitivity in the 2.4 GHz band reaching 0.9 indicates that slight noise here can significantly reduce the communication quality), and the low-sensitivity area (such as the sensitivity <0.1 in the frequency band below 10 MHz) reflects that the device has strong tolerance to the noise in this range. This data provides a basis for the frequency band priority sorting and tolerance threshold setting for subsequent impedance optimization.

[0111] It should be noted that based on the sensitivity distribution curve, the system calculates the deviation between the measured impedance characteristics and the theoretical ideal value to generate an impedance deviation index. This index reflects the current impedance matching defect in percentage form. For example, a deviation of 30% in the 100 MHz frequency band indicates that the parasitic inductance here causes too high impedance in the high-frequency return path. The index includes two types of key parameters: frequency-domain deviation (impedance deviation in a specific frequency band) and time-domain stability (the drift amount of impedance with temperature / load fluctuations), and the two jointly evaluate the actual performance of the circuit design's noise suppression ability.

[0112] It should be noted that by integrating the impedance deviation index and the equipment operating condition constraints, the system generates target impedance parameters through multi-parameter optimization. This parameter defines the impedance range that meets the electromagnetic compatibility requirements. For example, in the sensitive frequency band (such as 500 MHz - 1 GHz), the impedance value is required to be lower than 5 ohms to suppress common-mode noise. The optimization process balances three factors: electrical performance (such as insertion loss and return loss), physical implementation feasibility (such as PCB trace width and stack-up structure), and cost constraints (such as the amount of high-frequency magnetic materials used). The final target impedance parameters will guide the design of the filter circuit. For example, by adjusting the impedance-frequency characteristics of the ferrite beads, the impedance matching degree in the key frequency band can be increased to more than 95%.

[0113] In summary, the present invention discloses a method for shielding and reducing noise of a high-voltage harness of a new energy vehicle. The method includes obtaining current fluctuation data, working state encoding, and environmental condition data; performing wavelet decomposition on the current fluctuation data to obtain current fluctuation characteristics; performing environmental analysis on the environmental condition data to obtain environmental state characteristics; inputting the current fluctuation characteristics, the working state encoding, and the environmental state characteristics into a preset current prediction model to obtain predicted current data; extracting noise according to the predicted current data to obtain high-frequency noise data; and performing impedance matching according to the high-frequency noise data to output target impedance parameters. In a specific implementation process, first, the current fluctuation data is collected in real time through a Hall sensor array, and the working state encoding is transmitted through the vehicle controller and the CAN bus. At the same time, the temperature and humidity sensors and the EMI detection module integrated in the BMS system are used to collect environmental condition data. After these multi-source heterogeneous data are processed by time synchronization, they provide a high-quality data basis for subsequent steps such as wavelet decomposition and environmental analysis.

[0114] For the extraction of current fluctuation characteristics, a method of basis function matching is adopted. Appropriate basis functions (such as Daubechies or Symlets series) are selected according to the signal characteristics, and then recursive decomposition is performed to obtain a set of subband coefficients. The subband time-domain signals are obtained by reconstructing the set of subband coefficients, and finally, the current fluctuation characteristics are extracted from the subband time-domain signals using entropy quantization technology. This process can effectively capture the energy distribution in different frequency bands, helping to more accurately identify potential problems and providing accurate inputs for the current prediction model. At the same time, the analysis of environmental condition data is also crucial. It includes not only the monitoring of electromagnetic interference frequencies, but also considerations of complex environmental factors such as determining the proportion of high-frequency energy based on spectral analysis and identifying whether the propagation path is radiation propagation or waveguide propagation through periodic detection, so as to adjust system parameters to adapt to different working environments and reduce the impact of external changes on system performance.

[0115] Furthermore, by integrating the current fluctuation characteristics, working state encoding, and environmental state characteristics into a pre-trained current prediction model, the changing trend of the actual current can be more accurately simulated. This not only improves the prediction accuracy but also better reflects the response characteristics of the system under different working conditions, providing a reliable basis for subsequent noise extraction. In the noise extraction stage, advanced signal processing techniques such as baseline correction, frequency-domain transformation, and high-frequency filtering are used to separate the high-frequency noise components from the predicted current data. This is particularly crucial for subsequent impedance matching because correct impedance matching can effectively suppress noise and ensure the stable operation of the system. Finally, during the impedance matching based on the high-frequency noise data, a noise map is obtained through spectral extraction, and the sensitivity distribution curve is calculated. Then, the impedance deviation index is determined, and finally, multi-parameter optimization is achieved to obtain the optimal target impedance parameters, thereby improving the shielding and noise reduction accuracy of the high-voltage wire harness of new energy vehicles.

[0116] Referring to Figure 2 , the second embodiment of the present invention provides a shielding and noise reduction system for the high-voltage wire harness of a new energy vehicle, including:

[0117] A data acquisition module for acquiring current fluctuation data, working state encoding, and environmental condition data;

[0118] A current feature module for performing wavelet decomposition on the current fluctuation data to obtain current fluctuation characteristics;

[0119] An environmental feature module for performing environmental analysis on the environmental condition data to obtain environmental state characteristics;

[0120] A current prediction module for inputting the current fluctuation characteristics, the working state encoding, and the environmental state characteristics into a preset current prediction model to obtain predicted current data;

[0121] A noise extraction module, configured to extract noise based on the predicted current data to obtain high-frequency noise data;

[0122] An impedance matching module, configured to perform impedance matching based on the high-frequency noise data and output target impedance parameters.

[0123] It should be noted that the shielding and noise reduction system for a high-voltage harness of a new energy vehicle provided in an embodiment of the present invention is used to execute all the process steps of the method for shielding and noise reduction of a high-voltage harness of a new energy vehicle in the above embodiment. The working principles and beneficial effects of the two correspond one by one, and thus will not be elaborated herein.

[0124] An embodiment of the present invention further provides an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a data acquisition program. When the processor executes the computer program, the steps in the above embodiments of the method for shielding and noise reduction of a high-voltage harness of a new energy vehicle are implemented, such as Figure 1 step S11 shown. Alternatively, when the processor executes the computer program, the functions of each module / unit in the above device embodiments are implemented, such as the data acquisition module.

[0125] Exemplarily, the computer program may be divided into one or more modules / units. The one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing specific functions, and these instruction segments are used to describe the execution process of the computer program in the electronic device.

[0126] The electronic device may be a computing device such as a desktop computer, a notebook, a palm computer, and a smart tablet. The electronic device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the above components are only examples of the electronic device, and do not constitute a limitation to the electronic device. It may include more or fewer components than the above, or combine some components, or different components. For example, the electronic device may further include input / output devices, network access devices, a bus, etc.

[0127] The so-called processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the electronic device, connecting various parts of the entire electronic device through various interfaces and circuits.

[0128] The memory can be used to store the computer programs and / or modules. The processor realizes various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and by calling the data stored in the memory. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store the operating system, application programs required for at least one function (such as the sound playback function, the image playback function, etc.); the data storage area can store the data created according to the use of the mobile phone (such as audio data, phone book, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disks, memory, plug-in hard disks, Smart Media Cards (SMCs), Secure Digital (SD) cards, Flash Cards, at least one magnetic disk storage device, flash memory device, or other volatile solid-state storage devices.

[0129] Among them, if the modules / units integrated in the electronic device are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above-described embodiment methods of the present invention, it can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice within the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0130] It should be noted that the device embodiments described above are merely illustrative. 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 to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the attached drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that they have a communication connection, which can be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0131] The above-described specific embodiments have further elaborated on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is only the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. It is particularly pointed out that for those skilled in the art, any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A shielding and noise reduction method for a high-voltage wire harness of a new energy vehicle, characterized in that: include: Obtain current fluctuation data, working status code and environmental condition data; Performing wavelet decomposition according to the current fluctuation data to obtain current fluctuation characteristics; Performing environmental analysis according to the environmental condition data to obtain environmental state characteristics; Inputting the current fluctuation characteristics, the working state code and the environmental state characteristics into a preset current prediction model to obtain predicted current data; Perform noise extraction according to the predicted current data to obtain high-frequency noise data; Perform impedance matching according to the high-frequency noise data and output target impedance parameters; Wherein, the working state coding adopts one-hot coding or numerical mapping to represent the equipment operation mode and control instructions; The step of performing environmental analysis according to the environmental condition data to obtain environmental state characteristics includes: Perform spectrum analysis according to the environmental condition data to obtain the proportion of high-frequency energy; When the high-frequency energy ratio is greater than a preset high-frequency threshold, determining that the electromagnetic propagation path is radiation propagation; Perform periodic detection according to the environmental condition data to obtain a propagation period; When the propagation period is an integer, it is determined that the electromagnetic propagation path is waveguide propagation; Extracting field strength according to the electromagnetic propagation path and the environmental condition data to obtain field strength data; Feature fusion is performed based on the electromagnetic propagation path and the field strength data to obtain environmental state features.

2. The shielding and noise reduction method for high-voltage wire harness of new energy vehicles according to claim 1 is characterized in that: The obtaining of current fluctuation data, working state code and environmental condition data includes: Get the electromagnetic interference frequency; When the electromagnetic interference frequency is less than a preset first sampling frequency, data acquisition is performed according to the first sampling frequency to obtain current fluctuation data; When the electromagnetic interference frequency is greater than or equal to the first sampling frequency, performing frequency updating according to the electromagnetic interference frequency to obtain a second sampling frequency; When the electromagnetic interference frequency is greater than the second sampling frequency, data acquisition is performed according to the second sampling frequency to obtain current fluctuation data; Get the car acceleration; When the acceleration of the vehicle is greater than a preset first working threshold, determining that the working state code is a preset first code; When the vehicle acceleration is less than or equal to the first working threshold and greater than a preset second working threshold, determining that the working state code is a preset second code; When the vehicle acceleration is less than or equal to the second working threshold, the working state code is determined to be a preset third code.

3. The shielding and noise reduction method for high-voltage wire harness of new energy vehicles according to claim 1 is characterized in that: The step of performing wavelet decomposition according to the current fluctuation data to obtain current fluctuation characteristics includes: Perform basis function matching according to the current fluctuation data to obtain a decomposed basis function; Recursively decomposing the decomposition basis function and the current fluctuation data to obtain a sub-band coefficient set; Reconstructing a signal according to the sub-band coefficient set to obtain a sub-band time domain signal; The entropy value is quantized according to the sub-band time domain signal to obtain the current fluctuation characteristics.

4. The shielding and noise reduction method for high-voltage wire harness of new energy vehicles according to claim 1 is characterized in that: The training process of the current prediction model includes: A current prediction model is constructed based on a long short-term memory neural network, where the input layer includes historical current characteristics, historical working conditions, and historical environmental characteristics; The output layer is the target current prediction value; In the iterative process, the mean square error between the predicted value and the true value is used as the loss function; Complete one round of iteration to update parameters; When it is detected that the loss function of the model meets the conditions or the number of training times reaches the set upper limit, the training is considered to be completed and the trained model is obtained.

5. The shielding and noise reduction method for high-voltage wire harness of new energy vehicles according to claim 1 is characterized in that: The step of extracting noise according to the predicted current data to obtain high-frequency noise data includes: Performing baseline correction according to the predicted current data to obtain residual waveform data; Perform frequency domain transformation according to the residual waveform data to obtain energy spectrum density distribution; Perform high-frequency filtering according to the energy spectrum density distribution to obtain high-frequency noise distribution; Time-frequency joint labeling is performed according to the high-frequency noise distribution to obtain high-frequency noise data.

6. The shielding and noise reduction method for high-voltage wire harness of new energy vehicles according to claim 1 is characterized in that: The performing impedance matching according to the high-frequency noise data and outputting target impedance parameters comprises: Performing spectrum extraction according to the high-frequency noise data to obtain a noise spectrum; Performing sensitivity calculation according to the noise spectrum to obtain a sensitivity distribution curve; Performing deviation calculation according to the sensitivity distribution curve to obtain an impedance deviation index; Multi-parameter optimization is performed according to the impedance deviation index to obtain the target impedance parameters.

7. A shielding and noise reduction system for high-voltage wiring harnesses of new energy vehicles, characterized in that: A shielding and noise reduction method for a high-voltage wire harness of a new energy vehicle according to any one of claims 1 to 6, comprising: A data acquisition module, used to acquire current fluctuation data, working status code and environmental condition data; A current characteristic module, used for performing wavelet decomposition according to the current fluctuation data to obtain current fluctuation characteristics; An environmental characteristic module, used to perform environmental analysis based on the environmental condition data to obtain environmental state characteristics; A current prediction module, used for inputting the current fluctuation characteristics, the working state code and the environmental state characteristics into a preset current prediction model to obtain predicted current data; A noise extraction module, used for performing noise extraction according to the predicted current data to obtain high-frequency noise data; The impedance matching module is used to perform impedance matching according to the high-frequency noise data and output target impedance parameters.

8. An electronic device, characterized in that: It comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the shielding and noise reduction method for the high-voltage wire harness of a new energy vehicle as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the shielding and noise reduction method for the high-voltage wire harness of a new energy vehicle as described in any one of claims 1 to 6.

Images