Rotating speed adjusting method of low-noise motor fan

Abstract

The invention relates to the technical field of motor fan control, and discloses a rotating speed adjusting method of a low-noise motor fan. According to the method, a fan rotating speed adjusting model is trained through a deep learning algorithm based on an attention mechanism, and historical in-vehicle environment data and air conditioner working state data are used as training samples. Acquiring in-vehicle temperature, humidity and air conditioner state data in real time, inputting the data into the model, outputting the target rotating speed of the fan, and automatically adjusting the fan to the rotating speed. Meanwhile, the adjustment condition of a passenger is recorded through a camera in front of the seat, and personalized fan rotating speed setting is achieved. In addition, an environment perception compensation system is further introduced, vehicle external environment data are obtained in real time, an environment compensation coefficient is added in the model, the model is dynamically updated by adopting a federal learning framework, and the generalization ability and the real-time performance of the model are improved. Under extreme environmental conditions, a safe operation mode is activated to ensure safe operation of the vehicle and comfort of passengers.

Description

Technical Field

[0001] The invention relates to the technical field of motor fan control, and in particular to a rotation speed regulation method for a low-noise motor fan. Background Art

[0002] In the automotive interior environment control system, the speed regulation of the motor fan is crucial to maintaining a comfortable temperature and humidity in the car. Traditional fan speed regulation methods usually rely on simple temperature sensors and fixed speed settings, and cannot be intelligently adjusted according to the actual environmental changes in the car and the personal preferences of passengers. With the development of intelligent technology, intelligent control systems based on deep learning algorithms are gradually applied to the automotive field. However, the existing intelligent control systems still have shortcomings in fan speed regulation, such as insufficient model training, low data utilization, and inability to perform personalized adjustments based on passenger preferences. Therefore, a low-noise motor fan control method is needed that can use the in-car environment data and passenger preferences for intelligent speed regulation. Summary of the invention

[0003] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to provide a method for adjusting the speed of a low-noise motor fan, so as to solve the problem in the prior art that the fan speed adjustment has a low degree of intelligence and cannot be automatically and accurately adjusted according to the actual environmental changes in the vehicle and the personal preferences of the passengers.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a method for adjusting the speed of a low-noise motor fan, the method comprising: Step S100: using historical in-vehicle temperature data, humidity data, and automobile air conditioning working status data as training samples, and obtaining a fan speed adjustment model through deep learning algorithm training based on an attention mechanism; Step S200: collecting the current temperature data, humidity data and working status data of the car air conditioner in real time, inputting the real-time data into the fan speed adjustment model, and outputting the fan target speed; Step S300: According to the fan target speed, automatically adjusting the speed of the low-noise motor fan to the target speed; Step S400: The camera in front of the seat records the current occupant and the adjustment status of the motor fan speed. When the same occupant occupies the seat again, the adjustment record is read and the motor fan speed is automatically adjusted to the preferred setting.

[0005] Preferably, in a possible implementation manner of the first aspect, the fan speed adjustment model structure includes: The input layer is used to receive historical in-vehicle temperature data, humidity data, and automobile air conditioning working status data; The embedding layer encodes the data of the input layer and converts it into a dense vector representation in a high-dimensional space; The attention mechanism layer assigns different weights according to the importance of the input data features; Hidden layer, which consists of multiple neural network layers, is used to extract deep features of input data and integrate information through nonlinear transformation; The output layer maps the data features extracted by the hidden layer to the fan target speed value and outputs the target speed.

[0006] Preferably, in a possible implementation of the first aspect, a dynamic convolution module is deployed in the hidden layer, and the convolution kernel parameters are dynamically adjusted according to the weight output by the attention mechanism layer, specifically including: In each neuron node of the hidden layer, a deformable convolution kernel is embedded, and its shape parameters are dynamically calculated and generated by the attention weight; The deformation amplitude of the convolution kernel is constrained by the gated recurrent unit to ensure the effectiveness of the convolution operation; The dynamic convolution output is fused with the standard convolution output using residual connection to form a hybrid feature expression.

[0007] Preferably, in a possible implementation manner of the first aspect, step S200 specifically includes: Through temperature sensors, humidity sensors and air conditioning status monitoring devices, the current temperature data, humidity data and working status data of the car air conditioner are captured in real time; Perform data preprocessing on the collected real-time data, including data cleaning and outlier detection; The pre-processed real-time data is input into the trained fan speed adjustment model. The model automatically calculates and outputs the fan target speed required in the current environment based on the input data. Verify the fan target speed output by the model to ensure that it is within the preset reasonable range.

[0008] Preferably, in a possible implementation manner of the first aspect, step S300 specifically includes: receiving a fan target speed value output by a fan speed regulation model, using the fan target speed value as a desired fan operating speed; Compare the actual speed of the current low-noise motor fan with the target speed to determine the direction and amplitude of speed adjustment; According to the comparison result, a speed adjustment instruction is sent to the low-noise motor fan, and the instruction includes a speed adjustment amount; The motor adjusts the actual speed of the fan to the target speed according to the received command; During the adjustment process, the actual speed of the fan is monitored in real time to ensure that it is within the preset error range with the target speed.

[0009] Preferably, in a possible implementation manner of the first aspect, step S400 specifically includes: The camera in front of the seat performs face recognition to detect and record the information of the person currently sitting in the seat; matching the detected person information with previously stored passenger adjustment records, the adjustment records including each passenger's preference setting for the motor fan speed; When it is recognized that the same seat occupant is sitting again, the corresponding preference settings are retrieved from the database; Use the fan target speed as the initial setting and adjust the speed of the low-noise motor fan according to your preference.

[0010] Preferably, in a possible implementation of the first aspect, when the face recognition algorithm recognizes a certain passenger for the first time, the facial features of the passenger are collected and recorded in a database, and the preference settings of the passenger are recorded; When the facial features of a certain passenger are identified to match the facial feature records in the database, the corresponding preference settings are retrieved from the database.

[0011] Preferably, in a possible implementation manner of the first aspect, the preference setting includes: Read the current fan target speed and set the current fan target speed as the initial reference; When the passenger gets on the vehicle for the first time, the fan speed adjusted by the passenger is read, a deviation between the adjusted fan speed and the fan target speed is calculated, and the deviation is used as the passenger's preference setting and recorded in the database; When a passenger takes multiple rides, the preference is set to the average of the multiple deviation values.

[0012] Preferably, in a possible implementation manner of the first aspect, the method further includes an environment perception compensation system: Obtain real-time external environment data of the vehicle through the Internet of Vehicles, including weather conditions, road congestion index and air pollution index; Add an environmental compensation coefficient calculation module to the fan speed adjustment model to perform multi-modal fusion of external environmental data and in-vehicle sensor data; The federated learning framework is used to dynamically update the fan speed adjustment model, and the adjustment data of the vehicle group is distributedly trained through edge computing devices; When extreme environmental conditions are detected, the safe operating mode is activated, limiting the fan speed adjustment range and synchronously adjusting the air conditioning compressor operating frequency.

[0013] The beneficial effects of the present invention are as follows: by training the fan speed adjustment model based on the deep learning algorithm of the attention mechanism, the changes of complex factors such as the temperature, humidity and air conditioning working status in the vehicle can be accurately captured, and the intelligent adjustment of the fan speed can be realized. The in-vehicle environmental data is collected in real time and input into the model, and the fan target speed that meets the current environment is output, thereby improving the accuracy and timeliness of the fan speed adjustment. The passenger adjustment status is recorded by the camera in front of the seat, and the same passenger is provided with personalized fan speed settings, which enhances the riding experience.

[0014] In addition, the present invention also introduces an environmental perception compensation system, which can obtain the vehicle's external environmental data in real time, and adds an environmental compensation coefficient calculation module to the model, so that the fan speed adjustment can better adapt to changes in the external environment. The federated learning framework is used to dynamically update the model, and edge computing devices are used for distributed training to improve the generalization ability and real-time performance of the model. Under extreme environmental conditions, the safe operation mode is activated, the fan speed adjustment range is limited, and the operating frequency of the air-conditioning compressor is adjusted to ensure the safe operation of the vehicle and the comfort of the passengers. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 A flow chart of a method for adjusting the speed of a low-noise motor fan is provided for the present application. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Embodiment 1: Figure 1 As shown, the present invention provides a method for adjusting the speed of a low-noise motor fan, the method comprising: Step S100: Using historical in-vehicle temperature data, humidity data, and automobile air-conditioning working status data as training samples, a fan speed adjustment model is obtained through training with a deep learning algorithm based on an attention mechanism.

[0019] Specifically, historical in-vehicle temperature data, humidity data, and automobile air conditioning working status data are obtained through the automobile control system. The fan speed adjustment model structure includes: an input layer for receiving historical in-vehicle temperature data, humidity data, and automobile air conditioning working status data; an embedding layer for encoding the data in the input layer and converting it into a dense vector representation in a high-dimensional space; an attention mechanism layer for assigning different weights according to the importance of the input data features; a hidden layer, which contains multiple neural network layers, for extracting deep features of the input data and integrating information through nonlinear transformations; and an output layer for mapping the data features extracted by the hidden layer to the fan target speed value and outputting the target speed.

[0020] A dynamic convolution module is deployed in the hidden layer to dynamically adjust the convolution kernel parameters according to the weight of the attention mechanism layer output. Specifically, a deformable convolution kernel is embedded in each neuron node in the hidden layer, and its shape parameters are dynamically calculated and generated by the attention weight; the deformation amplitude of the convolution kernel is constrained by the gated recurrent unit to ensure the effectiveness of the convolution operation; the dynamic convolution output is fused with the standard convolution output using a residual connection to form a hybrid feature expression.

[0021] In this embodiment, historical in-vehicle temperature data, humidity data and automobile air conditioning working status data are used as training samples, and a prediction model is constructed based on an improved attention mechanism deep learning framework, which includes five parts: data input layer, feature embedding layer, attention weighting layer, dynamic convolution hidden layer and speed output layer. The in-vehicle environmental parameters collected in real time by the automobile control system are transmitted to the model input via the CAN bus, and the target fan speed command is generated after multi-layer feature conversion. The specific implementation process is as follows: The training samples are derived from the historical operation data collected by the vehicle sensors, including temperature sensor readings, humidity sensor signals, and air conditioning compressor working status codes in the form of time series. The sliding window method is used to construct the sample sequence in the data preprocessing stage. The window length is set to 20 time steps and the step length is set to 5 seconds. Each sample contains the environmental parameter change characteristics for 100 consecutive seconds. The input layer is designed as a multi-channel data structure, corresponding to three independent channels of temperature, humidity, and air conditioning status, and zero-mean normalization is used to eliminate dimensional differences.

[0022] The embedding layer uses a learnable encoding method to achieve data space mapping. A 32-dimensional embedding vector is designed for continuous variables such as temperature and humidity, and a 16-dimensional one-hot encoding extension is used for discrete variables such as air conditioning working status. The embedding process introduces a feature interaction mechanism, sets cross-channel association parameters in the embedding matrix, and establishes combined expressions such as temperature-humidity correlation features and air conditioning mode-temperature coupling features through matrix dot product operations. The encoded feature vector is promoted to a unified dimensional space through a fully connected network to form a fused feature tensor with a shape of [batch_size, seq_len, 64].

[0023] The attention mechanism layer adopts a multi-head scaling dot product attention architecture, with 4 independent attention heads working in parallel. The query matrix Q, key matrix K, and value matrix V of each attention head are extracted from the input features through linear transformation, where the Q matrix dimension is set to 256, and the K and V matrix dimensions are 128. The temperature coefficient adjustment mechanism is introduced into the attention weight calculation, which dynamically adjusts the exponential scaling amplitude of the softmax function through learnable parameters to enhance the response sensitivity to sudden environmental changes. After the weighted feature vector is normalized by the layer, it is residually connected with the original input to form a context vector with feature selection enhancement.

[0024] The dynamic convolution hidden layer uses a gated deformable convolution architecture to achieve feature extraction, and integrates the attention guidance mechanism on the basis of the standard convolutional neural network. Each hidden layer contains 32 deformable convolution kernels, whose spatial deformation parameters are dynamically generated by the attention weight matrix. The convolution kernel offset generator is constructed using a two-layer fully connected network. The first layer maps the 64-dimensional attention features to a 128-dimensional intermediate space, and the second layer outputs a coordinate offset that is times the square of the convolution kernel size. The gating mechanism generates a spatial mask through a sigmoid function and constrains the offset to ensure that the effective receptive field of the convolution operation does not exceed the range of 7x7 pixels. After the convolution output features are compressed by a depth-separable convolution, they are channel-weighted fused with the standard convolution path output, and the fusion coefficients are dynamically calculated by a gated recurrent unit.

[0025] The residual connection module adopts a cross-layer feature reuse strategy and introduces a skip connection path after each group of dynamic convolution operations. An adaptive weighting mechanism is designed in the feature fusion stage to perform a nonlinear combination of the original features and convolution outputs through a learnable parameter matrix. The combined weight is generated by the L2 norm value of the current feature map through a sigmoid function.

[0026] The output layer is designed as a multi-scale prediction structure, which includes an instant speed prediction branch and a trend prediction branch. The instant branch uses a fully connected layer to directly map the final hidden feature to the speed value, and the trend branch extracts the speed change gradient of the next 5 time steps through a time convolution network. The dual-branch output is weighted and summed in the channel dimension, and the weight parameter is dynamically adjusted according to the current air conditioning working mode: the instant branch weight is increased to 0.7 in the rapid cooling mode, and the trend branch weight is increased to 0.6 in the energy-saving mode. The final output layer constrains the predicted value to the [0,1] interval through the hyperbolic tangent function, and matches the actual fan speed range after linear transformation.

[0027] The model training adopts a phased optimization strategy. In the initial stage, the deformation parameters of the dynamic convolution module are frozen, and 100 rounds of pre-training are performed with the standard convolutional network. In the second stage, the parameters of the gated unit are unfrozen, and the cosine annealing learning rate scheduling algorithm is used for 50 rounds of fine-tuning, and the basic learning rate is set to 3e-4. In the final stage, all-parameter joint training is enabled, and the dynamic gradient clipping technology is introduced. The upper limit of the gradient norm is set to 0.5 to prevent parameter oscillation. The loss function is designed in a composite form, including the mean square error of speed prediction, the smoothness constraint of trend prediction, and the regularization term of convolution kernel deformation, and the weight ratio is set to 5:2:1. The data enhancement strategy adds random mask processing, randomly masking local fragments in the input sequence with a probability of 20%, to improve the model's fault tolerance to sensor failures.

[0028] Step S200: collecting the current temperature data, humidity data and working status data of the automobile air conditioner in real time, inputting the real-time data into the fan speed adjustment model, and outputting the fan target speed.

[0029] Specifically, temperature sensors, humidity sensors, and air conditioning status monitoring devices are arranged in the car. These sensors and monitoring devices capture the current temperature data, humidity data, and working status data of the car air conditioner in real time. These raw data often contain noise and outliers, so data preprocessing is required before inputting the data into the model.

[0030] In the data preprocessing stage, data cleaning is first performed to remove invalid or missing values ​​caused by sensor failure or data transmission errors. Next, outlier detection is performed to identify and process values ​​that significantly deviate from the normal range to ensure that the data input into the model is accurate and reliable.

[0031] In this embodiment, data cleaning is performed first. First, the df.dropna method is used to remove rows containing NaN values ​​to ensure the accuracy of subsequent analysis. Next, outlier detection is performed. For temperature data, the normal range is set to 0 to 40 degrees, and values ​​outside this range are considered outliers. For humidity data, we set the normal range to 5% to 90%. Similarly, any value outside this range is considered an outlier. For detected outliers, we choose to replace them with the upper and lower limits of the corresponding range to ensure the rationality and consistency of the data.

[0032] The format and dimension of the preprocessed real-time data already match the input requirements of the fan speed regulation model, so it can be directly input into the trained fan speed regulation model. Inside the model, the embedding layer converts the data into a dense vector representation in a high-dimensional space. Subsequently, the attention mechanism layer assigns different weights according to the importance of the input data features, emphasizing those features that are more critical to fan speed regulation. The hidden layer further extracts deep features and integrates information through nonlinear transformations. Finally, the output layer maps these features to fan target speed values ​​and outputs them.

[0033] In order to ensure that the fan target speed output by the model meets the actual needs and is within a reasonable range, it is verified, including comparison with the preset speed range, and backtesting with historical data to evaluate the accuracy and stability of the model output.

[0034] In this embodiment, the speed range is first set to 0 to 2000 revolutions per minute. After the model outputs the fan target speed based on the real-time in-vehicle environmental data (temperature, humidity, air conditioning status), it is compared with the preset range to ensure that the output is reasonable and meets actual needs. Secondly, backtesting is performed using historical data, including past in-vehicle environmental data and corresponding fan speed records. The historical data is preprocessed and input into the model to obtain the target speed, which is compared with the actual speed. The accuracy of the model is evaluated through indicators such as error rate and accuracy.

[0035] Step S300: According to the fan target speed, the speed of the low-noise motor fan is automatically adjusted to the target speed.

[0036] Specifically, a fan target speed value output by a fan speed regulation model is received and used as the desired fan operating speed; the actual speed of the current low-noise motor fan is compared with the target speed to determine the direction and amplitude of the speed regulation; based on the comparison result, a speed regulation instruction is sent to the low-noise motor fan, the instruction including the speed adjustment amount; the motor regulates the actual speed of the fan to the target speed based on the received instruction; during the regulation process, the actual speed of the fan is monitored in real time to ensure that it is within a preset error range with the target speed.

[0037] In this embodiment, after receiving the fan target speed value (set to N rpm) output by the fan speed adjustment model, the system obtains the actual speed of the current low-noise motor fan (set to M rpm). M is compared with N to determine the direction and amplitude of the speed adjustment. If M is less than N, the system generates an acceleration instruction, which includes the adjustment amount required to increase the speed to N; if M is greater than N, a deceleration instruction is generated, which includes the adjustment amount required to reduce the speed to N.

[0038] The motor gradually adjusts the actual speed of the fan through the internal speed regulation mechanism according to the received instructions. During the adjustment process, the system monitors the actual speed of the fan in real time to ensure that the error between it and the target speed N remains within the preset range of ±10 rpm.

[0039] Step S400: The camera in front of the seat records the current occupant and the adjustment status of the motor fan speed. When the same occupant occupies the seat again, the adjustment record is read and the motor fan speed of the occupant is automatically adjusted to the preferred setting.

[0040] Specifically, facial recognition is performed through a camera in front of the seat to detect and record the information of the person currently occupying the seat; the detected person information is matched with the previously stored passenger adjustment record, which includes each passenger's preference settings for the motor fan speed; when it is recognized that the same person is sitting in the seat again, the corresponding preference settings are retrieved from the database; the fan target speed is used as the initial setting, and the speed of the low-noise motor fan is adjusted according to the preference settings.

[0041] When the face recognition algorithm recognizes a passenger for the first time, it collects the passenger's facial features and records them in the database, and also records the passenger's preference settings. When the facial features of a passenger are identified and match the facial feature records in the database, the corresponding preference settings are retrieved from the database.

[0042] The preference setting step includes reading the current fan target speed and setting the current fan target speed as the initial reference; when the passenger takes the vehicle for the first time, reading the fan speed adjusted by the passenger, calculating the deviation between the adjusted fan speed and the fan target speed, and recording the deviation as the passenger's preference setting in the database; when the passenger takes the vehicle multiple times, the preference setting is the average of multiple deviation values. In this embodiment, the camera in front of the seat uses an IR monocular camera and adopts face recognition technology to capture and identify the person currently sitting in the seat in real time. Through the face recognition algorithm, the system can detect and record the information of the person currently sitting in the seat, including the passenger's facial features, identity information, etc.

[0043] This embodiment uses the ArcFace algorithm. The ArcFace algorithm is a face recognition method based on deep learning. It extracts facial features by constructing a deep neural network and uses these features for face recognition and comparison. When the system detects a passenger, its facial features are matched with the passenger adjustment records stored in the database. These adjustment records include each passenger's previous preference settings for the motor fan speed. These preference settings are generated by manually adjusting the fan speed and saving the adjustment results when the passenger was on the vehicle before.

[0044] In order to realize the automatic reading and adjustment of passenger preferences, a passenger preference database is established. The database is used to store each passenger's facial features, identity information, and their preferences for the motor fan speed. When the system recognizes that the facial features of a passenger match the facial feature records in the database, the corresponding preference settings will be automatically retrieved from the database. When the same seat is recognized to be occupied again, the system will automatically read their preferences and use the fan target speed as the initial setting. Then, the speed of the low-noise motor fan is adjusted according to the preference settings.

[0045] The adjustment process of the preference setting is as follows: First, the system reads the current fan target speed and sets it as the initial benchmark. Then, when the passenger gets on the vehicle for the first time, the system reads the adjusted fan speed based on the passenger's adjustment operation. Next, the system calculates the deviation between the adjusted fan speed and the fan target speed and records this deviation value as the passenger's preference setting in the database. This deviation value reflects the passenger's personalized demand for fan speed.

[0046] When a passenger takes a ride multiple times, the system will collect statistics and average their preferences. Specifically, the system calculates the average of multiple deviations and uses this average as the final preference for the passenger. When adjusting the fan speed, the system monitors the actual fan speed in real time and makes fine adjustments based on the preference settings to ensure that the fan speed is consistent with the passenger's preference settings. At the same time, the system will also record the results of each adjustment and the feedback from the passengers in order to continuously optimize and improve the fan speed adjustment algorithm.

[0047] Embodiment 2: The present invention provides a method for adjusting the speed of a low-noise motor fan, and the method also includes an environmental perception compensation system.

[0048] Specifically, the vehicle's external environmental data, including weather conditions, road congestion index and air pollution index, is obtained in real time through the Internet of Vehicles. An environmental compensation coefficient calculation module is added to the fan speed adjustment model to perform multimodal fusion of external environmental data and in-vehicle sensor data. The fan speed adjustment model is dynamically updated using a federated learning framework, and the adjustment data of the vehicle group is distributedly trained through edge computing devices. When extreme environmental conditions are detected, the safe operation mode is activated to limit the fan speed adjustment range and synchronously adjust the air-conditioning compressor operating frequency.

[0049] In this embodiment, the Internet of Vehicles system accesses the meteorological service cloud platform and the traffic management center database through the 5G-V2X communication protocol, and collects real-time detailed environmental parameters such as precipitation intensity, wind speed and direction, and PM2.5 concentration at the location of the vehicle. The environmental compensation coefficient calculation module uses the attention mechanism to extract features from heterogeneous data, normalizes discrete weather codes and continuous pollution indexes, and generates a dynamic compensation factor in the range of 0.8-1.2. This factor is combined with the operating condition data provided by the in-vehicle temperature sensor and the battery thermal management system through a weighted fusion algorithm to form a composite heat dissipation demand index, which serves as the reference input for fan speed adjustment.

[0050] The federated learning framework is used to continuously optimize model parameters. Vehicle terminals upload local training gradients instead of original data through edge computing nodes, which not only ensures privacy security but also improves model generalization capabilities. Differential privacy technology is introduced during training to add Gaussian noise to the gradient to ensure convergence stability in data heterogeneity scenarios. After the model of the edge server aggregates the vehicle group, it pushes the optimized adjustment strategy to the fleet through OTA, forming a closed-loop learning mechanism of "local training-global sharing".

[0051] When the system detects that environmental parameters exceed preset thresholds, such as encountering sandstorms or extremely high temperatures, the safe operation mode immediately activates a dual protection mechanism. On the one hand, the PID controller limits the fan speed fluctuation to On the other hand, it establishes a collaborative channel with the air conditioning control system to adjust the compressor operating frequency in proportion to the actual fan speed to maintain the air pressure balance inside and outside the cabin. In this mode, the heat dissipation efficiency of key components is prioritized, and the heat dissipation efficiency is maximized within a limited speed range by dynamically adjusting the angle of the air guide plate at the air conditioning outlet.

[0052] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for adjusting the speed of a low-noise motor fan, characterized in that: The method comprises: Step S100: using historical in-vehicle temperature data, humidity data, and automobile air conditioning working status data as training samples, and obtaining a fan speed adjustment model through deep learning algorithm training based on an attention mechanism; Step S200: collecting the current temperature data, humidity data and working status data of the car air conditioner in real time, inputting the real-time data into the fan speed adjustment model, and outputting the fan target speed; Step S300: According to the fan target speed, automatically adjusting the speed of the low-noise motor fan to the target speed; Step S400: The camera in front of the seat records the current occupant and the adjustment status of the motor fan speed. When the same occupant occupies the seat again, the adjustment record is read and the motor fan speed is automatically adjusted to the preferred setting.

2. The speed regulation method according to claim 1, characterized in that: The fan speed regulation model structure includes: The input layer is used to receive historical in-vehicle temperature data, humidity data, and automobile air conditioning working status data; The embedding layer encodes the data of the input layer and converts it into a dense vector representation in a high-dimensional space; The attention mechanism layer assigns different weights according to the importance of the input data features; Hidden layer, which consists of multiple neural network layers, is used to extract deep features of input data and integrate information through nonlinear transformation; The output layer maps the data features extracted by the hidden layer to the fan target speed value and outputs the target speed.

3. The speed regulation method according to claim 2, characterized in that: A dynamic convolution module is deployed in the hidden layer to dynamically adjust the convolution kernel parameters according to the weight output by the attention mechanism layer, specifically including: In each neuron node of the hidden layer, a deformable convolution kernel is embedded, and its shape parameters are dynamically calculated and generated by the attention weight; The deformation amplitude of the convolution kernel is constrained by the gated recurrent unit to ensure the effectiveness of the convolution operation; The dynamic convolution output is fused with the standard convolution output using residual connection to form a hybrid feature expression.

4. The speed regulation method according to claim 1, characterized in that: The step S200 specifically includes: Through temperature sensors, humidity sensors and air conditioning status monitoring devices, the current temperature data, humidity data and working status data of the car air conditioner are captured in real time; Perform data preprocessing on the collected real-time data, including data cleaning and outlier detection; The pre-processed real-time data is input into the trained fan speed adjustment model. The model automatically calculates and outputs the fan target speed required in the current environment based on the input data. Verify the fan target speed output by the model to ensure that it is within the preset reasonable range.

5. The speed regulation method according to claim 1, characterized in that: The step S300 specifically includes: receiving a fan target speed value output by a fan speed regulation model, using the fan target speed value as a desired fan operating speed; Compare the actual speed of the current low-noise motor fan with the target speed to determine the direction and amplitude of speed adjustment; According to the comparison result, a speed adjustment instruction is sent to the low-noise motor fan, and the instruction includes a speed adjustment amount; The motor adjusts the actual speed of the fan to the target speed according to the received command; During the adjustment process, the actual speed of the fan is monitored in real time to ensure that it is within the preset error range with the target speed.

6. The speed regulation method according to claim 1, characterized in that: The step S400 specifically includes: The camera in front of the seat performs face recognition to detect and record the information of the person currently sitting in the seat; matching the detected person information with previously stored passenger adjustment records, the adjustment records including each passenger's preference setting for the motor fan speed; When it is recognized that the same seat occupant is sitting again, the corresponding preference settings are retrieved from the database; Use the fan target speed as the initial setting and adjust the speed of the low-noise motor fan according to your preference.

7. The speed regulation method according to claim 6, characterized in that: When the face recognition algorithm recognizes a passenger for the first time, it collects the passenger's facial features and records them in the database, and also records the passenger's preference settings; When the facial features of a certain passenger are identified to match the facial feature records in the database, the corresponding preference settings are retrieved from the database.

8. The speed regulation method according to claim 7, characterized in that: The preferences include: Read the current fan target speed and set the current fan target speed as the initial reference; When the passenger gets on the vehicle for the first time, the fan speed adjusted by the passenger is read, a deviation between the adjusted fan speed and the fan target speed is calculated, and the deviation is used as the passenger's preference setting and recorded in the database; When a passenger takes multiple rides, the preference is set to the average of the multiple deviation values.

9. The speed regulation method according to claim 1, characterized in that: The method also includes an environmental perception compensation system: Obtain real-time vehicle external environment data through the Internet of Vehicles, including weather conditions, road congestion index and air pollution index; Add an environmental compensation coefficient calculation module to the fan speed adjustment model to perform multi-modal fusion of external environmental data and in-vehicle sensor data; The federated learning framework is used to dynamically update the fan speed adjustment model, and the adjustment data of the vehicle group is distributedly trained through edge computing devices; When extreme environmental conditions are detected, the safe operating mode is activated, limiting the fan speed adjustment range and synchronously adjusting the air conditioning compressor operating frequency.

Images