Comprehensive experimental device for refrigeration and heat exchange of air conditioner
By building a temperature change prediction model and anomaly detection module in the comprehensive experimental device of air conditioner refrigeration and heat exchange, the problem of temperature change prediction and abnormal detection of performance tests in the existing technology is solved, and more accurate temperature prediction and experimental safety are achieved. promote.
Patent Information
- Application Number
- CN202410971951.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing performance testing devices of automobile air conditioning systems lack deep mining and intelligent analysis capabilities, it is difficult to accurately predict temperature change trends, and lack effective abnormal detection methods, which affects the safety and effectiveness of the experiment.
A comprehensive experimental device for air conditioner cooling and heat exchange is designed, including machine learning layer, data acquisition layer, data processing layer and monitoring and feedback layer. Machine learning technology is used to build a temperature change prediction model, collect and process data in real time, and perform abnormal detection. Provide user interface display and operation functions.
By constructing a temperature change prediction model, the temperature change trend of the air-conditioning pipeline system can be accurately predicted, and abnormal conditions can be discovered and dealt with in a timely manner, which improves the safety of the experiment and the effectiveness of the data.
Smart Images

Figure CN118913738A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of automobile air-conditioning performance testing, in particular to an air-conditioning refrigeration and heat exchange comprehensive experimental device. Background Art
[0002] With the rapid development of the modern automobile industry, the performance optimization and intelligent control of automobile air-conditioning systems have become key factors in improving driving comfort. Most existing experimental devices rely on traditional data recording and analysis methods, lacking the ability to deeply explore and intelligently analyze the operating parameters of key components such as compressors, condensers, and evaporators. This makes it difficult to accurately predict the temperature change trend of the air-conditioning system under specific working conditions during the experiment, and cannot provide strong data support for system design and optimization.
[0003] Moreover, most of the existing air conditioning, refrigeration and heat exchange comprehensive experimental devices remain at the stage of real-time monitoring and test data display, without a unified abnormality detection judgment standard and a lack of more intelligent abnormality detection means, which makes it difficult to timely discover and deal with abnormal conditions that may occur in the air conditioning system during the experiment, affecting the safety and effectiveness of the experiment. Therefore, we propose an air conditioning, refrigeration and heat exchange comprehensive experimental device. Summary of the invention
[0004] The purpose of the present invention is to provide an air conditioning, refrigeration and heat exchange comprehensive experimental device to solve the problems raised in the above background technology.
[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an air conditioning refrigeration and heat exchange comprehensive experimental device, comprising a machine learning layer, a data acquisition layer, a data processing layer, and a monitoring and feedback layer;
[0006] The machine learning layer includes a model training module;
[0007] The model training module is used to construct a temperature change prediction model using machine learning technology; the temperature change prediction model predicts the temperature change trend of the air outlet of the automobile air conditioning pipeline system in the future period of time based on the production parameters of the compressor, condenser, and evaporator, as well as the ambient temperature, wind speed setting, and temperature setting;
[0008] The data acquisition layer includes a sensor module;
[0009] The sensor module is used to collect the temperature at the air outlet of the tested air conditioning pipeline system in real time;
[0010] The data processing layer includes a data preprocessing module, a data sending module and a data storage module;
[0011] The data preprocessing module is used to clean and denoise the raw data collected by the sensor module, the data sending module is used to transmit the preprocessed data to the monitoring and feedback layer in real time, and the data storage module is used to store the preprocessed data persistently;
[0012] The monitoring and feedback layer includes a real-time monitoring module, a model calling module, an anomaly detection module and a user interface module;
[0013] The real-time monitoring module is used to receive data transmitted by the data sending module, and plot the real-time received temperature acquisition data into a curve that changes over time; the model calling module is used to call the temperature change prediction model, and the specific steps are:
[0014] i) Obtain model input parameters, including the production parameters of the compressor, condenser, and evaporator of the current automotive air conditioning system, as well as the ambient temperature, wind speed setting, and temperature setting,
[0015] ii) passing the input parameters into the temperature change prediction model,
[0016] iii) receiving a predicted temperature change curve at an air outlet of the air conditioning pipeline system under current input parameters;
[0017] The abnormality detection module is used to compare the difference between the temperature change curve output by the temperature change prediction model and the temperature change curve output by the real-time monitoring module, and generate a difference curve by calculating the difference between the two curve data. The specific steps are:
[0018] i) obtaining the predicted temperature change curve data output by the model calling module and the real-time temperature change curve data output by the real-time monitoring module;
[0019] ii) align the timestamps of the two curve data to make the data at each time point correspond to each other;
[0020] iii) calculating the difference between the predicted temperature and the real-time temperature at the corresponding time point, and generating a difference curve;
[0021] iv) judging whether each point on the difference curve exceeds the normal range according to a preset threshold value, and if so, marking it as an abnormal point;
[0022] v) transmitting the difference curve and abnormal point information to the user interface module;
[0023] The user interface module is used to display real-time monitoring data and model prediction data, as well as the difference data and abnormal order information between the two, and provide interface operation functions to users.
[0024] Preferably, the specific steps of constructing and training the temperature change prediction model by the model training module include:
[0025] Step 1, collecting historical data, wherein the historical data is time series data of normal air outlet temperature changes of the automobile air conditioning system under specific working parameters, wherein the specific working parameters include production parameters of the compressor, condenser, and evaporator, as well as ambient temperature, wind speed setting, and temperature setting;
[0026] Step 2: Clean the collected historical data to handle missing values and outliers;
[0027] Step 3: Standardize the cleaned feature data to form a training data set;
[0028] Step 4: Divide the training data set into a training set and a validation set;
[0029] Step 5: Select the long short-term memory network LSTM model to build a temperature change prediction model and initialize the parameters of the model;
[0030] Step 6: Use the training set data to iteratively train the LSTM model. During the training process, regularly use the validation set to evaluate the model performance until the model performance meets the requirements.
[0031] Step 7. After training is completed, save the weights and biases of the LSTM model and deploy the model for subsequent temperature change prediction.
[0032] Preferably, in the above step 2, for missing values, a linear interpolation algorithm is used to fill them; for outliers, a threshold judgment method is used to regard data points that exceed a preset threshold range as outliers and replace them with the mean of their neighboring points.
[0033] Preferably, in step 3, the Z-score standardization method is used to subtract the mean of each numerical data and divide it by its standard deviation so that the processed data conforms to the standard normal distribution. The formula for the standardization process is: Among them, X is the original data, μ is the mean of the original data, σ is the standard deviation of the original data, and Z is the standardized data.
[0034] Preferably, the parameters for initializing the LSTM model in step 5 specifically include:
[0035] i) Initializing the structural parameters of the LSTM model, including the feature dimension of the input layer, the number of hidden layers and the number of hidden units in each layer, and the number of nodes in the output layer, wherein the feature dimension of the input layer matches the number of extracted power signal features, and the number of nodes in the output layer is equal to the number of predicted optimal sampling point positions;
[0036] ii) Initialize the model weights and biases using small random numbers;
[0037] iii) Setting the activation function of the LSTM model, including setting the Sigmoid function for the gating mechanism and setting the Tanh function for the calculation of candidate memory cells and hidden states;
[0038] iv) Configure the optimization algorithm and set the learning rate parameters.
[0039] Preferably, in step 6, the LSTM model is iteratively trained using the training set data. In each iteration, a batch of input data is sent to the LSTM model, and the memory unit and hidden state are updated by calculating the values of the input gate, the forget gate and the output gate, and finally the predicted output sequence is obtained; the loss function value is calculated according to the predicted output and the true label; the gradient of the loss function with respect to the model parameters is calculated by the back propagation algorithm; the model weights and biases are updated according to the gradient using the optimization algorithm; during the training process, the validation set is regularly used to evaluate the model performance, and the learning rate is adjusted or the training is stopped early as needed to prevent overfitting.
[0040] Preferably, the mean square error MSE algorithm is used to evaluate the performance of the temperature change prediction model. The specific algorithm is:
[0041]
[0042] Where n is the number of data points, Y pred,i is the temperature value predicted by the model, Y actual,i is the actual measured temperature value; the prediction accuracy of the model is evaluated according to the MSE value. The smaller the MSE value, the higher the prediction accuracy of the model.
[0043] Preferably, the user interface module includes a real-time monitoring data display area, a model prediction data display area and a difference data display area;
[0044] The real-time temperature data transmitted by the receiving data transmission module is plotted into a curve changing with time and displayed in the real-time monitoring data display area;
[0045] The predicted temperature change curve is obtained by calling the temperature change prediction model and displayed in the model prediction data display area;
[0046] By comparing the real-time monitoring data with the model prediction data, the difference between the two is calculated, and the difference data is plotted as a curve and displayed in the difference data display area.
[0047] Preferably, the user interface module provides a user input interface to receive model parameters input by the user, provides control buttons to allow the user to start or stop real-time monitoring and prediction functions, and provides a data export interface to allow the user to export displayed data.
[0048] Compared with the prior art, the present invention has the following beneficial effects:
[0049] Through the model training module in the machine learning layer, a temperature change prediction model is constructed using machine learning technology. This model can comprehensively consider the production parameters of the compressor, condenser, and evaporator, as well as multiple factors such as ambient temperature, wind speed setting, and temperature setting, and accurately predict the temperature change trend of the air outlet of the automobile air conditioning pipeline system in the future. The abnormal detection module in the monitoring and feedback layer can compare the difference between the temperature change curve output by the temperature change prediction model and the temperature change curve output by the real-time monitoring module in real time. By calculating the difference and generating the difference curve, combined with the preset threshold to judge the abnormal point, the abnormal conditions that may occur in the air conditioning system can be discovered and handled in a timely manner, effectively improving the safety of the experiment and the effectiveness of the data. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] Figure 1 It is the overall structure diagram of the present invention;
[0051] Figure 2 Diagram of the steps for building and training a temperature change prediction model. DETAILED DESCRIPTION
[0052] 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.
[0053] See also Figure 1-2 The present invention provides a technical solution: an air conditioning, refrigeration and heat exchange comprehensive experimental device, including a machine learning layer, a data acquisition layer, a data processing layer, and a monitoring and feedback layer.
[0054] The machine learning layer includes a model training module, whose core function is to use machine learning technology to build a temperature change prediction model. This model can predict the temperature change trend of the air outlet of the automobile air-conditioning pipeline system in the future based on the production parameters of the compressor, condenser, and evaporator, as well as multi-dimensional inputs such as ambient temperature, wind speed setting, and temperature setting, so that the temperature change trend under specific working conditions can be accurately predicted.
[0055] The data acquisition layer includes a sensor module, which is responsible for real-time acquisition of the temperature at the air outlet of the air conditioning pipeline system under test.
[0056] The data processing layer includes a data preprocessing module, a data transmission module, and a data storage module. The data preprocessing module cleans and denoises the raw data collected by the sensor module to ensure the accuracy and availability of the data. The data transmission module is responsible for transmitting the preprocessed data to the monitoring and feedback layer in real time to detect abnormal data in real time. The data storage module stores the preprocessed data persistently for subsequent historical data queries.
[0057] The monitoring and feedback layer includes real-time monitoring module, model calling module, anomaly detection module and user interface module.
[0058] The real-time monitoring module receives the data transmitted by the data sending module, and plots the real-time received temperature acquisition data into a curve that changes over time, so that the user can intuitively understand the real-time operating status of the air-conditioning system.
[0059] The model calling module is responsible for calling the temperature change prediction model. The specific steps include obtaining the model input parameters, passing the input parameters into the temperature change prediction model, and receiving the predicted temperature change curve at the air outlet of the air-conditioning pipe system under the current input parameters, so that users can predict the temperature change trend of the air-conditioning system under specific working conditions.
[0060] The anomaly detection module compares the difference between the temperature change curve output by the temperature change prediction model and the temperature change curve output by the real-time monitoring module, and generates a difference curve by calculating the difference between the two curve data. The specific steps include obtaining the predicted and real-time temperature change curve data, aligning the timestamps of the two curve data, calculating the difference between the predicted temperature and the real-time temperature at the corresponding time point, judging whether each point on the difference curve exceeds the normal range according to the preset threshold, and transmitting the difference curve and abnormal point information to the user interface module, so that users can promptly discover and deal with possible abnormal conditions of the air-conditioning system, improving the safety and effectiveness of the experiment.
[0061] The user interface module displays real-time monitoring data and model prediction data, as well as the difference data and abnormal point information between the two, and provides users with interface operation functions, allowing users to easily view and analyze data, and promptly discover and handle abnormal conditions, thereby improving the convenience of the experiment and user experience.
[0062] The present invention will be further described below in conjunction with Examples 1 to 3:
[0063] Embodiment 1:
[0064] The Long Short-Term Memory (LSTM) model is suitable for processing time series data because it can capture long-term dependencies in the data. In the temperature prediction of the automotive air-conditioning system, the change in the outlet temperature is often affected by multiple working parameters over a period of time in the past. The LSTM model can effectively learn and utilize this historical information through its unique memory unit and gating mechanism, thereby more accurately predicting future temperature change trends. The temperature change of the automotive air-conditioning system is affected by many factors, and there may be complex nonlinear relationships between these factors. As a deep learning model, the LSTM model has a strong representation learning ability and can handle such complex nonlinear relationships. It can effectively learn and simulate complex dynamic behaviors in actual systems through a multi-layer network structure and nonlinear activation functions.
[0065] The model training module uses the LSTM model to build a temperature change prediction model. The specific steps include:
[0066] Step 1, collecting historical data, wherein the historical data is time series data of normal air outlet temperature changes of the automobile air conditioning system under specific working parameters, wherein the specific working parameters include production parameters of the compressor, condenser, and evaporator, as well as ambient temperature, wind speed setting, and temperature setting;
[0067] Step 2: Clean the collected historical data to deal with missing values and outliers. For missing values, use linear interpolation algorithm to fill them. For outliers, use threshold judgment method to regard data points that exceed the preset threshold range as outliers and replace them with the mean of their neighboring points.
[0068] Using the Z-score standardization method, each numerical data is subtracted from its mean and divided by its standard deviation, so that the processed data conforms to the standard normal distribution. The formula for standardization is: Among them, X is the original data, μ is the mean of the original data, σ is the standard deviation of the original data, and Z is the standardized data.
[0069] Step 3: Standardize the cleaned feature data to eliminate the dimensional differences between different features so that the model can treat each feature more equally during the training process. For the time series data of the outlet temperature change of the automobile air-conditioning system, it is regarded as a feature sequence, and each feature in the sequence is standardized. After the processing is completed, the standardized feature sequence is paired with the corresponding timestamp label to form a training data set. Each data in the training data set contains a feature sequence and a corresponding timestamp label, which is used for matching input and output during subsequent model training.
[0070] Step 4: Use random partitioning to divide the data in the training data set into a training set and a validation set in a certain ratio (such as 7:3 or 8:2), and index and mark the training set and validation set so that they can be used correctly in the subsequent model training process. In this way, it can ensure that the model can fully learn the characteristics of the data during the training process and can effectively predict and evaluate unknown data.
[0071] Step 5: Select the long short-term memory network LSTM model to build a temperature change prediction model and initialize the parameters of the model; the specific steps include:
[0072] i) Initialize the structural parameters of the LSTM model, including determining the feature dimension of the input layer, which matches the number of extracted power signal features to ensure that the model can receive and process all relevant input information. At the same time, it is necessary to set the number of hidden layers and the number of hidden units in each layer. The choice of these parameters will affect the complexity and learning ability of the model. Finally, determine the number of nodes in the output layer, which is equal to the number of predicted optimal sampling point locations, to ensure that the model can output accurate prediction results.
[0073] ii) Initialize the model weights and biases. In order to avoid the problem of gradient vanishing or exploding in the early stage of model training, small random numbers are used to initialize the model weights and biases. This ensures that the model can stably update parameters during training and gradually learn the inherent laws of the data.
[0074] iii) Set the activation function of the LSTM model. In order to build an effective gating mechanism, the Sigmoid function is selected as the activation function, which can map the input value to the interval (0,1), thereby achieving effective control of information. In order to calculate the candidate memory cells and hidden states, the Tanh function is selected as the activation function, which can map the input value to the interval (-1,1) and has good gradient characteristics.
[0075] iv) Configure the optimization algorithm and set the learning rate parameters. To optimize the parameters of the LSTM model, select a suitable optimization algorithm, such as the Adam optimization algorithm, and set an appropriate learning rate parameter, which will help the model converge to the optimal solution quickly and accurately during the training process.
[0076] Step 6: Use the training set data to iteratively train the LSTM model. During the training process, regularly use the validation set to evaluate the model performance until the model performance meets the requirements. The specific steps include:
[0077] a) In each iteration, a batch of input data is fed into the LSTM model. This batch of data contains the time series data of the outlet temperature change of the automobile air conditioning system under specific working parameters.
[0078] b) By calculating the values of the input gate, forget gate, and output gate, the LSTM model can update the memory cells and hidden states and finally obtain the predicted output sequence. The model uses its internal complex structure and gating mechanism to capture the time series characteristics in the data.
[0079] c) Calculate the loss function value based on the predicted output and the true label. The loss function is used to measure the difference between the model's predicted results and the true results. The model parameters will be optimized by minimizing the loss function.
[0080] d) The gradient of the loss function with respect to the model parameters is calculated through the back-propagation algorithm. This step is the core of the optimization algorithm. It passes the gradient information of the loss function to the parameters of the model so that the parameters can be updated in subsequent iterations.
[0081] e) Use an optimization algorithm to update the model weights and biases based on the gradient. After obtaining the gradient information, use the optimization algorithm to update the model weights and biases so that the model can better fit the training data in subsequent iterations.
[0082] f) During the training process, regularly use the validation set to evaluate the model performance. The validation set is a data set independent of the training set that is used to evaluate the model's generalization ability on unknown data. By regularly evaluating the performance of the model on the validation set, you can monitor the model's training process and adjust the learning rate or stop training early as needed to prevent overfitting.
[0083] Step 7. After training, the LSTM model has learned the inherent rules in the data and reflected them in the weights and biases of the model. Save the weights and biases of the LSTM model so that the same model can be used in subsequent temperature change predictions.
[0084] The temperature change prediction model effectively utilizes LSTM's powerful modeling capabilities for time series data, realizes accurate prediction of the temperature changes at the air outlet of the automobile air-conditioning system, and improves the intelligence level of the system.
[0085] Below is a code example of building and training a temperature change prediction model using Python as the programming language:
[0086] import numpy as np
[0087] import pandas as pd
[0088] from keras.models import Sequential
[0089] from keras.layers import LSTM,Dense
[0090] from sklearn.model_selection import train_test_split
[0091] from sklearn.preprocessing import StandardScaler
[0092] #DataFrame containing historical data, represented by df
[0093] #df = pd.read_csv('your_data.csv')#Read data
[0094] #Step 2: Data cleaning (here is an example of cleaning method)
[0095] #df=df.dropna()#Remove missing values
[0096] #df=df[(df['temperature']>-100)&(df['temperature']<100)]#Remove abnormal temperature values
[0097] #Step 3: Data Standardization
[0098] scaler = StandardScaler()
[0099] scaled_features=scaler.fit_transform(df[['compressor_param','condenser_param','evaporator_param','ambient_temperature','fan_speed','temperature_set']])
[0100] #The prediction target is the temperature at the next time point
[0101] X = scaled_features[:-1,:]
[0102] y = scaled_features[1:,0] # Temperature prediction
[0103] #Step 4: Divide the training set and validation set
[0104] X_train,X_val,y_train,y_val=train_test_split(X,y,test_size=0.2,random_state=42)
[0105] #Reshape the data into the format expected by LSTM [samples, time steps, features]
[0106] X_train=np.reshape(X_train,(X_train.shape[0],1,X_train.shape[1]))
[0107] X_val=np.reshape(X_val,(X_val.shape[0],1,X_val.shape[1]))
[0108] #Step 5: Build LSTM model
[0109] model=Sequential()
[0110] model.add(LSTM(50,input_shape=(1,X_train.shape[2])))
[0111] model.add(Dense(1))
[0112] model.compile(optimizer='adam',loss='mean_squared_error')
[0113] #Step 6: Train the model
[0114] model.fit(X_train,y_train,epochs=100,batch_size=32,validation_data=(X_val,y_val),verbose=1)
[0115] #Step 7: Save the model
[0116] model.save('temperature_prediction_model.h5')
[0117] In the above code, NumPy is used for data reshaping and basic mathematical operations; Pandas is used to process raw historical data, including data cleaning and preprocessing; Keras is used to build the LSTM model, including defining the model structure, compiling the model, and training the model; the model_selection module provides data partitioning functions, such as dividing data into training sets and test sets; the preprocessing module provides data preprocessing functions, such as data standardization and normalization.
[0118] Embodiment 2:
[0119] The mean square error (MSE) algorithm is used to evaluate the performance of the temperature change prediction model. The specific algorithm is:
[0120]
[0121] Where n is the number of data points, Y pred,i is the temperature value predicted by the model, Y actual,i is the actual measured temperature value; the prediction accuracy of the model is evaluated according to the MSE value. The smaller the MSE value, the higher the prediction accuracy of the model. For each sample in the test data set, the square of the difference between the temperature change predicted by the model and the actual temperature change is calculated; the square differences of all samples are summed and then divided by the total number of samples to obtain the MSE value. The smaller the MSE value, the closer the model's prediction result is to the actual data, and the better the model performance. By comparing the MSE values of different models, the prediction performance of each model can be intuitively evaluated, so as to select the optimal model.
[0122] In-depth analysis of the MSE value can further understand the prediction performance of the model under different working parameters, as well as the overall stability and generalization ability of the model. Combined with actual application scenarios, the MSE value can be reasonably interpreted to provide support for the subsequent optimization and application of the model.
[0123] Embodiment 3:
[0124] The anomaly detection module is used to compare the difference between the temperature change curve output by the temperature change prediction model and the temperature change curve output by the real-time monitoring module, and generate a difference curve by calculating the difference between the two curve data, and then detect and mark abnormal points. The following are the specific implementation steps of the anomaly detection module:
[0125] i) Data acquisition: The anomaly detection module obtains the predicted temperature change curve data from the model calling module. These data are predicted by the temperature change prediction model based on the input parameters of the current automobile air-conditioning system (such as the production parameters of the compressor, condenser, and evaporator, ambient temperature, wind speed setting, and temperature setting, etc.). At the same time, the real-time temperature change curve data is obtained from the real-time monitoring module. These data are obtained by drawing the real-time received temperature acquisition data into a curve that changes over time.
[0126] ii) Data alignment: Since the predicted temperature change curve and the real-time temperature change curve may have inconsistent timestamps, the two curve data need to be timestamp aligned to ensure that the data at each time point is consistent. Through interpolation or interception, the two curves have corresponding temperature data at each time point.
[0127] iii) Difference calculation: After the two curves are aligned, the difference between the predicted temperature and the real-time temperature at the corresponding time point is calculated to generate a difference curve. The difference curve reflects the difference between the predicted temperature and the real-time temperature.
[0128] iv) Abnormality judgment: judge whether each point on the difference curve exceeds the normal range according to the preset threshold. If the difference exceeds the set threshold, the point is considered abnormal and marked as an abnormal point. The preset threshold can be adjusted according to the needs of the actual application scenario to ensure the accuracy and sensitivity of abnormality detection.
[0129] v) Result transmission: The generated difference curve and the marked abnormal point information are transmitted to the user interface module so that the user can intuitively understand the abnormal situation of temperature change and take corresponding measures to deal with it in time.
[0130] The user interface module is an important part of the interaction between the system and the user. It is responsible for displaying real-time monitoring data, model prediction data and the difference data between the two, and provides user input interface, control buttons and data export interface.
[0131] The data display area implementation includes:
[0132] Real-time monitoring data display area: Receive the real-time temperature data transmitted by the data sending module, use the chart library (such as ECharts, Highcharts, etc.) to draw these data into a curve changing over time, and update and display them in the monitoring data display area in real time, so that users can intuitively understand the temperature changes of the current automobile air-conditioning system.
[0133] Model prediction data display area: By calling the temperature change prediction model, the predicted temperature change curve data is obtained. The data is also plotted into a curve using the chart library and displayed in the prediction data display area so that users can view the model's prediction results for future temperature changes.
[0134] Difference data display area: Compare the real-time monitoring data with the model prediction data, calculate the temperature difference between the two at each time point, plot the difference data into a curve, and display it in the difference data display area so that users can intuitively understand the accuracy of the model prediction and the difference between the actual temperature and the predicted temperature.
[0135] The user operation functions provided include:
[0136] Provide user input interfaces, such as text boxes and drop-down menus, to receive model parameters input by users, such as ambient temperature, wind speed setting, temperature setting, etc. These parameters will be used as input conditions for model prediction.
[0137] Control buttons such as "Start Monitoring", "Stop Monitoring", "Start Prediction", "Stop Prediction", etc. are provided to allow users to start or stop real-time monitoring and prediction functions as needed, thereby improving the flexibility of the system and user experience.
[0138] Provide a data export interface, such as an export button or an export option in the right-click menu, to allow users to export displayed data into files in formats such as Excel and CSV, making it easier for users to conduct subsequent data analysis and processing.
[0139] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.
[0140] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. An air conditioning refrigeration and heat exchange comprehensive experimental device, characterized in that: It includes machine learning layer, data collection layer, data processing layer, monitoring and feedback layer; The machine learning layer includes a model training module; The model training module is used to construct a temperature change prediction model using machine learning technology; the temperature change prediction model predicts the temperature change trend of the air outlet of the automobile air conditioning pipeline system in the future period of time based on the production parameters of the compressor, condenser, and evaporator, as well as the ambient temperature, wind speed setting, and temperature setting; The data acquisition layer includes a sensor module; The sensor module is used to collect the temperature at the air outlet of the tested air conditioning pipeline system in real time; The data processing layer includes a data preprocessing module, a data sending module and a data storage module; The data preprocessing module is used to clean and denoise the raw data collected by the sensor module, the data sending module is used to transmit the preprocessed data to the monitoring and feedback layer in real time, and the data storage module is used to store the preprocessed data persistently; The monitoring and feedback layer includes a real-time monitoring module, a model calling module, an anomaly detection module and a user interface module; The real-time monitoring module is used to receive data transmitted by the data sending module, and plot the real-time received temperature acquisition data into a curve that changes over time; the model calling module is used to call the temperature change prediction model, and the specific steps are: i) Obtain model input parameters, including the production parameters of the compressor, condenser, and evaporator of the current automotive air conditioning system, as well as the ambient temperature, wind speed setting, and temperature setting, ii) passing the input parameters into the temperature change prediction model, iii) receiving a predicted temperature change curve at an air outlet of the air conditioning pipeline system under current input parameters; The abnormality detection module is used to compare the difference between the temperature change curve output by the temperature change prediction model and the temperature change curve output by the real-time monitoring module, and generate a difference curve by calculating the difference between the two curve data. The specific steps are: i) obtaining the predicted temperature change curve data output by the model calling module and the real-time temperature change curve data output by the real-time monitoring module; ii) align the timestamps of the two curve data to make the data at each time point correspond to each other; iii) calculating the difference between the predicted temperature and the real-time temperature at the corresponding time point, and generating a difference curve; iv) judging whether each point on the difference curve exceeds the normal range according to a preset threshold value, and if so, marking it as an abnormal point; v) transmitting the difference curve and abnormal point information to the user interface module; The user interface module is used to display real-time monitoring data and model prediction data, as well as the difference data and abnormal order information between the two, and provide interface operation functions to users.
2. The air conditioning, refrigeration and heat exchange comprehensive experimental device according to claim 1 is characterized in that: The specific steps of constructing and training the temperature change prediction model by the model training module include: Step 1, collecting historical data, wherein the historical data is time series data of normal air outlet temperature changes of the automobile air conditioning system under specific working parameters, wherein the specific working parameters include production parameters of the compressor, condenser, and evaporator, as well as ambient temperature, wind speed setting, and temperature setting; Step 2: Clean the collected historical data to handle missing values and outliers; Step 3: Standardize the cleaned feature data to form a training data set; Step 4: Divide the training data set into a training set and a validation set; Step 5: Select the long short-term memory network LSTM model to build a temperature change prediction model and initialize the parameters of the model; Step 6: Use the training set data to iteratively train the LSTM model. During the training process, regularly use the validation set to evaluate the model performance until the model performance meets the requirements. Step 7. After training is completed, save the weights and biases of the LSTM model and deploy the model for subsequent temperature change prediction.
3. The air conditioning, refrigeration and heat exchange comprehensive experimental device according to claim 2 is characterized by: In the above step 2, for missing values, linear interpolation algorithm is used to fill them; for outliers, threshold judgment method is used to regard data points beyond the preset threshold range as outliers and replace them with the mean of their neighboring points.
4. The air conditioning, refrigeration and heat exchange comprehensive experimental device according to claim 3 is characterized in that: In step 3, the Z-score standardization method is used to subtract the mean of each numerical data and divide it by its standard deviation, so that the processed data conforms to the standard normal distribution. The formula for standardization is: Among them, X is the original data, μ is the mean of the original data, σ is the standard deviation of the original data, and Z is the standardized data.
5. The air conditioning, refrigeration and heat exchange comprehensive experimental device according to claim 4 is characterized in that: The parameters for initializing the LSTM model in step 5 specifically include: i) Initializing the structural parameters of the LSTM model, including the feature dimension of the input layer, the number of hidden layers and the number of hidden units in each layer, and the number of nodes in the output layer, wherein the feature dimension of the input layer matches the number of extracted power signal features, and the number of nodes in the output layer is equal to the number of predicted optimal sampling point positions; ii) Initialize the model weights and biases using small random numbers; iii) Setting the activation function of the LSTM model, including setting the Sigmoid function for the gating mechanism and setting the Tanh function for the calculation of candidate memory cells and hidden states; iv) Configure the optimization algorithm and set the learning rate parameters.
6. The air conditioning, refrigeration and heat exchange comprehensive experimental device according to claim 5 is characterized in that: In step 6, the training set data is used to iteratively train the LSTM model. In each iteration, a batch of input data is sent to the LSTM model, and the memory cells and hidden states are updated by calculating the values of the input gate, forget gate, and output gate to finally obtain the predicted output sequence; the loss function value is calculated based on the predicted output and the true label; the gradient of the loss function with respect to the model parameters is calculated through the back propagation algorithm; the model weights and biases are updated according to the gradient using the optimization algorithm; during the training process, the validation set is regularly used to evaluate the model performance, and the learning rate is adjusted or the training is stopped early as needed to prevent overfitting.
7. The air conditioning, refrigeration and heat exchange comprehensive experimental device according to claim 6 is characterized in that: The mean square error (MSE) algorithm is used to evaluate the performance of the temperature change prediction model. The specific algorithm is: Where n is the number of data points, Y pred,i is the temperature value predicted by the model, Y actual,i is the actual measured temperature value; the prediction accuracy of the model is evaluated according to the MSE value. The smaller the MSE value, the higher the prediction accuracy of the model.
8. The air conditioning, refrigeration and heat exchange comprehensive experimental device according to claim 7 is characterized in that: The user interface module includes a real-time monitoring data display area, a model prediction data display area and a difference data display area; The real-time temperature data transmitted by the receiving data transmission module is plotted into a curve changing with time and displayed in the real-time monitoring data display area; The predicted temperature change curve is obtained by calling the temperature change prediction model and displayed in the model prediction data display area; By comparing the real-time monitoring data with the model prediction data, the difference between the two is calculated, and the difference data is plotted as a curve and displayed in the difference data display area.
9. The air conditioning, refrigeration and heat exchange comprehensive experimental device according to claim 8, characterized in that: The user interface module provides a user input interface to receive model parameters input by the user, provides control buttons to allow the user to start or stop real-time monitoring and prediction functions, and provides a data export interface to allow the user to export displayed data.
Citation Information
Patent Citations
Method for fault identification and early warning of air conditioners in passenger compartments of motor train unit based on BP neural network
CN106600047A
Load prediction method and device for building air conditioning system and electronic equipment
CN116227725A
Beidou-based grounding wire hitching state analysis and abnormity early warning method
CN117761577A
Industrial internet data real-time prediction system and method
CN118260562A
Estimation calculation method for cruising mileage and energy consumption of ship
CN118313247A
Cited By
Cabin air conditioner prediction method based on bidirectional interaction Mama with independent channels
CN119719679A
Microorganism data visualization chart generation system and method based on model parameters
CN120388622A
