Data indicator prediction method and system for automobile air conditioning

Abstract

The present invention provides a data indicator prediction method and system for an automotive air conditioner, relating to the field of data processing technology. In the present invention, for each of multiple historical time values ​​included in a target time period, multiple historical air conditioning performance indicator parameters of the target automotive air conditioner at that historical time value are extracted. An updated target data indicator prediction neural network is used to analyze and process the multiple historical air conditioning performance indicator parameters corresponding to the multiple historical time values ​​to output a target air conditioning performance prediction result corresponding to the target automotive air conditioner. Multiple target air conditioning performance indicator parameters of the target automotive air conditioner at the current time are extracted, and based on the multiple target air conditioning performance indicator parameters and the target air conditioning performance prediction result, a corresponding target air conditioning performance fusion result is obtained. Based on the above, the reliability of data indicator prediction for automotive air conditioners can be improved to a certain extent.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and system for predicting data indicators of automobile air conditioners. Background Art

[0002] Automotive air conditioners typically have numerous performance indicators, such as the actual cooling EER, the actual heating EER, the noise level of the indoor unit, and the noise level of the outdoor unit. However, existing techniques typically analyze these indicators separately, or simply combine their averages to predict the final comprehensive performance indicator. This results in poor reliability in predicting automotive air conditioner data indicators. Summary of the Invention

[0003] In view of this, an object of the present invention is to provide a method and system for predicting data indicators of automobile air conditioners, so as to improve the reliability of the prediction of data indicators of automobile air conditioners to a certain extent.

[0004] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0005] A data indicator prediction method for automobile air conditioning, comprising:

[0006] For each of the multiple historical time values ​​included in the target time period, extracting multiple historical air-conditioning performance index parameters of the target automobile air-conditioner at the historical time value;

[0007] Using the updated target data indicator prediction neural network, a plurality of historical air conditioning performance indicator parameters corresponding to the plurality of historical time values ​​are analyzed and processed to output a target air conditioning performance prediction result corresponding to the target automobile air conditioner, the target air conditioning performance prediction result being used to reflect the degree of excellence of the comprehensive performance of the target automobile air conditioner;

[0008] A plurality of target air-conditioning performance index parameters of the target automobile air-conditioner at the current time are extracted, and a target air-conditioning performance fusion result corresponding to the target automobile air-conditioner is obtained based on the plurality of target air-conditioning performance index parameters and the target air-conditioning performance prediction result.

[0009] In some preferred embodiments, in the above-mentioned data indicator prediction method for an automobile air conditioner, the step of extracting, for each of the multiple historical time values ​​included in the target time period, multiple historical air conditioning performance indicator parameters of the target automobile air conditioner at the historical time value includes:

[0010] Obtaining the most recently analyzed historical target air conditioning performance fusion result, and then updating the pre-configured reference time period duration based on the historical target air conditioning performance fusion result to output a corresponding target time period duration;

[0011] At the current time, a corresponding target time period is determined based on the target time period duration, wherein the target time period duration is equal to the target time period duration, and the target time period end point is the current time;

[0012] For each historical time value included in the target time period, a plurality of historical air-conditioning performance index parameters of the target automobile air-conditioner at the historical time value are extracted. The target time period includes a plurality of historical time values.

[0013] In some preferred embodiments, in the above-mentioned data indicator prediction method for an automobile air conditioner, the step of using the updated target data indicator prediction neural network to analyze and process the multiple historical air conditioning performance indicator parameters corresponding to the multiple historical time values ​​to output a target air conditioning performance prediction result corresponding to the target automobile air conditioner includes:

[0014] Matrixing the multiple historical air conditioning performance index parameters corresponding to the multiple historical time values ​​to form a corresponding historical index parameter distribution matrix, wherein a row of the historical air conditioning performance index parameters is a value of a historical air conditioning performance index parameter at the multiple historical time values, and a column of the historical air conditioning performance index parameters is a value of multiple historical air conditioning performance index parameters at one historical time value;

[0015] The target data indicator prediction neural network is used to analyze and process the historical indicator parameter distribution matrix to output a target air-conditioning performance prediction result corresponding to the target automobile air-conditioning.

[0016] In some preferred embodiments, in the above-mentioned data indicator prediction method for automobile air conditioners, the step of using a target data indicator prediction neural network to analyze and process the historical indicator parameter distribution matrix to output a target air conditioning performance prediction result corresponding to the target automobile air conditioner includes:

[0017] Using the first matrix overall information mining model included in the updated target data indicator prediction neural network to perform a matrix overall information mining operation on the historical indicator parameter distribution matrix to output a matrix overall information to-be-processed vector corresponding to the historical indicator parameter distribution matrix;

[0018] Performing a key data extraction operation on the historical indicator parameter distribution matrix, and outputting a key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix;

[0019] Using the first matrix key information mining model included in the target data indicator prediction neural network to perform a matrix key information mining operation on the key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix, so as to output a matrix key information to-be-processed vector corresponding to the historical indicator parameter distribution matrix;

[0020] Analyze and output the target matrix information to be processed vector corresponding to the historical indicator parameter distribution matrix based on the matrix overall information to be processed vector and the corresponding matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix;

[0021] respectively calculating the vector similarity between the target matrix information to-be-processed vector corresponding to the historical indicator parameter distribution matrix and each of the pre-configured multiple target matrix information reference vectors, and then, based on the vector similarity between each target matrix information reference vector and the target matrix information to-be-processed vector, fusing the target air-conditioning performance labeling result corresponding to each target matrix information reference vector to form a target air-conditioning performance prediction result corresponding to the target automobile air-conditioning;

[0022] The first matrix key information mining model is formed by updating model parameters of the matrix key information mining model to be updated based on the first matrix overall information mining model and the extracted exemplary data combination, wherein the exemplary data combination includes an exemplary standard indicator parameter distribution matrix, an exemplary first indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix, and an exemplary second indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix;

[0023] In the process of updating the model parameters of the matrix critical information mining model to be updated, the first matrix overall information mining model is used to perform matrix overall information mining operations on the matrix overall information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix. The matrix critical information mining model to be updated is used to perform matrix critical information mining operations on the matrix critical information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix. The first matrix critical information mining model updates the model parameters of the matrix critical information mining model to be updated based on the key information mining error and information mining joint error output by analysis. The key information mining error is analyzed and output according to the matrix critical information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix respectively. The information mining joint error is analyzed and output according to the matrix overall information to be processed vectors and the matrix critical information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix respectively.

[0024] In some preferred embodiments, in the above-mentioned data indicator prediction method for automobile air conditioners, the step of performing a key data extraction operation on the historical indicator parameter distribution matrix and outputting a key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix includes:

[0025] Performing a numerical mapping operation on each parameter value of the historical air-conditioning performance index parameter included in the historical index parameter distribution matrix;

[0026] Replacing parameter values ​​of key historical air-conditioning performance index parameters in the historical index parameter distribution matrix after the numerical mapping operation with pre-configured key parameter values, and replacing parameter values ​​of non-key historical air-conditioning performance index parameters in the historical index parameter distribution matrix after the numerical mapping operation with pre-configured non-key parameter values, so as to form a key historical index parameter replacement distribution matrix corresponding to the historical index parameter distribution matrix, wherein the parameter values ​​corresponding to the key historical air-conditioning performance index parameters after the numerical mapping operation are not less than the pre-configured parameter comparison values, and the parameter values ​​corresponding to the non-key historical air-conditioning performance index parameters after the numerical mapping operation are less than the parameter comparison value;

[0027] A matrix fusion operation is performed on the key historical indicator parameter replacement distribution matrix and the historical indicator parameter distribution matrix to output a key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix.

[0028] In some preferred embodiments, in the above-mentioned data indicator prediction method for automobile air conditioners, the data indicator prediction method further includes:

[0029] Extracting the exemplary data combination, loading the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix included in the exemplary data combination into the first matrix overall information mining model to complete the matrix overall information mining operation, and outputting the standard matrix overall information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the correlation matrix overall information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the non-correlation matrix overall information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix;

[0030] Analyzing the key historical indicator parameter distribution matrices corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, respectively, loading the key historical indicator parameter distribution matrices corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, respectively, into the matrix key information mining model to be updated to complete the matrix key information mining operation, and outputting the matrix key information to-be-processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, respectively;

[0031] According to the standard matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the corresponding standard target matrix information to be processed vector is analyzed and outputted; then, according to the correlation matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, the corresponding correlation target matrix information to be processed vector is analyzed and outputted; then, according to the non-correlation matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix, the corresponding non-correlation target matrix information to be processed vector is analyzed and outputted;

[0032] Analyze and output the corresponding key information mining error based on the matrix key information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the matrix key information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the matrix key information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix;

[0033] Analyze and output the corresponding information mining joint error based on the standard target matrix information to be processed vector, the relevant target matrix information to be processed vector, and the non-relevant target matrix information to be processed vector;

[0034] The model parameters of the matrix critical information mining model to be updated are updated according to the critical information mining error and the information mining joint error to form an updated matrix critical information mining model, and the first matrix critical information mining model is obtained according to the updated matrix critical information mining model.

[0035] In some preferred embodiments, in the above-mentioned data indicator prediction method for automobile air conditioners, the steps of analyzing and outputting the corresponding standard target matrix information to be processed vector based on the standard matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, then analyzing and outputting the corresponding relevant target matrix information to be processed vector based on the relevant matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, and then analyzing and outputting the corresponding non-relevant target matrix information to be processed vector based on the non-relevant matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix, include:

[0036] Loading the matrix key information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the matrix key information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the matrix key information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix into the vector dimensionality reduction model to be updated to complete the vector dimensionality reduction operation, and outputting the reduced dimensionality matrix key information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the reduced dimensionality matrix key information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the reduced dimensionality matrix key information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix;

[0037] Based on the standard matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the corresponding standard target matrix information to be processed vector is analyzed and output; then, based on the relevant matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, the corresponding relevant target matrix information to be processed vector is analyzed and output; and, based on the non-relevant matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix, the corresponding non-relevant target matrix information to be processed vector is analyzed and output.

[0038] In some preferred embodiments, in the above-mentioned data indicator prediction method for automobile air conditioners, the step of analyzing and outputting the target matrix information to be processed vector corresponding to the historical indicator parameter distribution matrix based on the matrix overall information to be processed vector and the corresponding matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix includes:

[0039] A first vector dimensionality reduction model is used to perform a vector dimensionality reduction operation on the matrix key information to-be-processed vector corresponding to the historical indicator parameter distribution matrix, and output a dimensionality reduction matrix key information to-be-processed vector corresponding to the historical indicator parameter distribution matrix, wherein the first vector dimensionality reduction model is formed by performing a model parameter update operation on the vector dimensionality reduction model to be updated based on the key information mining error and / or the information mining joint error;

[0040] A vector aggregation operation is performed on the matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix to output the target matrix information to be processed vector corresponding to the historical indicator parameter distribution matrix.

[0041] In some preferred embodiments, in the above-mentioned data indicator prediction method for an automobile air conditioner, the step of extracting multiple target air conditioning performance indicator parameters of the target automobile air conditioner at the current time, and obtaining a target air conditioning performance fusion result corresponding to the target automobile air conditioner based on the multiple target air conditioning performance indicator parameters and the target air conditioning performance prediction result includes:

[0042] Extracting multiple target air conditioning performance index parameters of the target automobile air conditioner at the current time, performing weighted fusion processing on the multiple target air conditioning performance index parameters to output a calculation result of the actual air conditioning performance corresponding to the target automobile air conditioner;

[0043] A weighted sum calculation is performed on the actual air-conditioning performance calculation result and the target air-conditioning performance prediction result to obtain a target air-conditioning performance fusion result corresponding to the target automobile air-conditioner.

[0044] An embodiment of the present invention also provides a data indicator prediction system for automobile air conditioners, including a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to execute the computer program to implement the above-mentioned data indicator prediction method for automobile air conditioners.

[0045] Embodiments of the present invention provide a data indicator prediction method and system for an automotive air conditioner. For each of multiple historical time values ​​included in a target time period, multiple historical air conditioning performance indicator parameters of the target automotive air conditioner at that historical time value are extracted. An updated target data indicator prediction neural network is used to analyze and process the multiple historical air conditioning performance indicator parameters corresponding to the multiple historical time values ​​to output a target air conditioning performance prediction result corresponding to the target automotive air conditioner. Multiple target air conditioning performance indicator parameters of the target automotive air conditioner at the current time are extracted, and a corresponding target air conditioning performance fusion result is obtained based on the multiple target air conditioning performance indicator parameters and the target air conditioning performance prediction result. Based on the foregoing, the target data indicator prediction neural network can be used to perform a comprehensive analysis of a large number of historical air conditioning performance indicator parameters to output a corresponding target air conditioning performance prediction result. The multiple current target air conditioning performance indicator parameters are then fused with the target air conditioning performance prediction result to obtain a corresponding target air conditioning performance fusion result. This provides a more comprehensive basis for analyzing (predicting) the data indicators, thereby improving the reliability of the automotive air conditioner data indicator prediction to a certain extent.

[0046] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a structural block diagram of a data indicator prediction system for automobile air conditioners provided by an embodiment of the present invention.

[0048] Figure 2 A flowchart illustrating the steps of a method for predicting data indicators of an automobile air conditioner provided by an embodiment of the present invention.

[0049] Figure 3 A schematic diagram of various modules included in a data indicator prediction device for an automobile air conditioner provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0051] Reference Figure 1 As shown, an embodiment of the present invention provides a data indicator prediction system for automobile air conditioners. The data indicator prediction system may include a memory and a processor.

[0052] Further, in some embodiments, the memory and processor are directly or indirectly electrically connected to enable data transmission or exchange. For example, the electrical connection can be achieved via one or more communication buses or signal lines. The memory can store at least one software function module (computer program) in the form of software or firmware. The processor can be configured to execute the executable computer program stored in the memory, thereby implementing the data indicator prediction method for an automotive air conditioner provided in an embodiment of the present invention.

[0053] To further supplement, in some embodiments, the memory may be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. The processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a system on a chip (SoC), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0054] As a further supplementary explanation, in some embodiments, the data indicator prediction system for automobile air conditioners may be a server with data processing capabilities (which may also include a communication unit).

[0055] Combine Figure 2 The embodiment of the present invention also provides a data index prediction method for automobile air conditioners, which can be applied to the above-mentioned data index prediction system for automobile air conditioners. The method steps defined in the process related to the data index prediction method for automobile air conditioners can be implemented by the data index prediction system. Figure 2 The specific process shown is explained in detail.

[0056] Step S110 , for each of the multiple historical time values ​​included in the target time period, extract multiple historical air-conditioning performance index parameters of the target automobile air-conditioner at the historical time value.

[0057] In an embodiment of the present invention, the data indicator prediction system can extract, for each of the multiple historical time values ​​included in the target time period, multiple historical air conditioning performance indicator parameters of the target automobile air conditioner at the historical time value (for example, the actual cooling energy efficiency ratio, the actual heating energy efficiency ratio, the noise value of the indoor unit, the noise value of the outdoor unit, etc.)

[0058] In step S120 , the updated target data index prediction neural network is used to analyze and process the multiple historical air conditioning performance index parameters corresponding to the multiple historical time values ​​to output a target air conditioning performance prediction result corresponding to the target automobile air conditioner.

[0059] In an embodiment of the present invention, the data indicator prediction system can utilize the updated target data indicator prediction neural network to analyze and process the various historical air conditioning performance indicator parameters corresponding to the multiple historical time values, thereby outputting a target air conditioning performance prediction result corresponding to the target vehicle air conditioner. The target air conditioning performance prediction result is used to reflect the overall performance excellence of the target vehicle air conditioner (i.e., the overall performance predicted based on historical data).

[0060] Step S130 , extracting a plurality of target air conditioning performance index parameters of the target automobile air conditioner at the current time, and obtaining a target air conditioning performance fusion result corresponding to the target automobile air conditioner based on the plurality of target air conditioning performance index parameters and the target air conditioning performance prediction result.

[0061] In an embodiment of the present invention, the data index prediction system can extract a plurality of target air-conditioning performance index parameters of the target automobile air-conditioner at the current time, and obtain the target air-conditioning performance fusion result corresponding to the target automobile air-conditioner (used to reflect the comprehensive performance) based on the plurality of target air-conditioning performance index parameters and the target air-conditioning performance prediction result.

[0062] Based on the above content, we can first use the target data indicator prediction neural network to conduct an overall analysis of a large number of historical air-conditioning performance indicator parameters to output the corresponding target air-conditioning performance prediction results, and then fuse the current multiple target air-conditioning performance indicator parameters and the target air-conditioning performance prediction results to obtain the corresponding target air-conditioning performance fusion results. In this way, the basis for analyzing (predicting) the data indicators can be more sufficient. Therefore, the reliability of the data indicator prediction of the automobile air-conditioning can be improved to a certain extent, so that the problem of insufficient prediction reliability in the existing technology can be improved.

[0063] Further supplementary explanation, in some embodiments, during the process of executing step S110, the following steps may be performed instead:

[0064] Obtaining the most recently analyzed historical target air conditioning performance fusion result, and then updating a pre-configured reference time period length based on the historical target air conditioning performance fusion result to output a corresponding target time period length (illustratively, the better the performance reflected by the historical target air conditioning performance fusion result, the smaller the updated target time period length);

[0065] At the current time, a corresponding target time period is determined based on the target time period duration, wherein the target time period duration is equal to the target time period duration, and the target time period end point is the current time;

[0066] For each historical time value included in the target time period (such as a historical time value formed every 1 or 2 hours, etc.), multiple historical air-conditioning performance index parameters of the target automobile air-conditioning at the historical time value are extracted, and the target time period includes multiple historical time values.

[0067] Further supplementary explanation, in some embodiments, during the step of using the updated target data indicator prediction neural network to analyze and process the multiple historical air conditioning performance indicator parameters corresponding to the multiple historical time values ​​to output the target air conditioning performance prediction result corresponding to the target automobile air conditioner, the following steps may be performed instead:

[0068] Matrix processing is performed on the multiple historical air conditioning performance index parameters corresponding to the multiple historical time values ​​to form a corresponding historical index parameter distribution matrix, where a row of the historical air conditioning performance index parameter represents the values ​​of one historical air conditioning performance index parameter at the multiple historical time values, and a column of the historical air conditioning performance index parameter represents the values ​​of the multiple historical air conditioning performance index parameters at one historical time value (illustratively, the row data and the column data may also be swapped; in addition, the multiple historical time values ​​may be sorted from early to late in time, and the multiple historical air conditioning performance index parameters may be randomly sorted or may be sorted in a configured order);

[0069] The target data indicator prediction neural network is used to analyze and process the historical indicator parameter distribution matrix to output a target air-conditioning performance prediction result corresponding to the target automobile air-conditioning.

[0070] Further supplementary explanation, in some embodiments, during the process of executing the step of using the target data indicator prediction neural network to analyze and process the historical indicator parameter distribution matrix to output the target air-conditioning performance prediction result corresponding to the target automobile air-conditioner, the following steps may be performed instead:

[0071] The first matrix overall information mining model included in the updated target data indicator prediction neural network is used to perform a matrix overall information mining operation on the historical indicator parameter distribution matrix to output a matrix overall information to-be-processed vector corresponding to the historical indicator parameter distribution matrix (the matrix overall information to-be-processed vector reflects the overall information of the historical indicator parameter distribution matrix; in addition, exemplary, the matrix overall information mining model to be updated is updated with the exemplary data combination to form the first matrix overall information mining model; the matrix overall information mining model to be updated is used to mine the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary first indicator parameter distribution matrix in the exemplary data combination. The matrix overall information to be processed vectors corresponding to the exemplary second indicator parameter distribution matrix can be used to determine corresponding information mining errors based on the matrix overall information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, and then perform a model parameter update operation on the matrix overall information mining model to be updated based on the information mining error to output an updated matrix overall information mining model to be updated, and then use the updated matrix overall information mining model to be updated as the first matrix overall information mining model, or, construct the first matrix overall information mining model based on the model parameters of the updated matrix overall information mining model to be updated);

[0072] Performing a key data extraction operation on the historical indicator parameter distribution matrix, and outputting a key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix;

[0073] A first matrix critical information mining model included in the target data indicator prediction neural network is used to perform a matrix critical information mining operation on the key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix to output a matrix critical information to-be-processed vector corresponding to the historical indicator parameter distribution matrix. The first matrix critical information mining model is formed by updating model parameters of the matrix critical information mining model to be updated based on the first matrix overall information mining model and the extracted exemplary data combination. The exemplary data combination includes an exemplary standard indicator parameter distribution matrix, an exemplary first indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix, and an exemplary second indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix (exemplarily, the matrix similarity between the exemplary standard indicator parameter distribution matrix and the exemplary first indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix can be greater than a matrix similarity reference value, and the matrix similarity between the exemplary standard indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix can be less than the matrix similarity reference value; in addition, when updating model parameters of the matrix critical information mining model to be updated In the process, the first matrix overall information mining model is used to perform matrix overall information mining operations on the matrix overall information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix, and the matrix key information mining model to be updated is used to perform matrix key information mining operations on the matrix key information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix, and the first matrix key information mining model updates the model parameters of the matrix key information mining model to be updated based on the key information mining error and information mining joint error output by the analysis, the key information mining error is output according to the analysis of the matrix key information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix respectively, and the information mining joint error is output according to the analysis of the matrix overall information to be processed vectors and the matrix key information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix respectively);

[0074] Analyze and output the target matrix information to be processed vector corresponding to the historical indicator parameter distribution matrix based on the matrix overall information to be processed vector and the corresponding matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix (exemplarily, the target data indicator prediction neural network can be used to analyze and output the target matrix information to be processed vector);

[0075] The vector similarities between the target matrix information to-be-processed vector corresponding to the historical indicator parameter distribution matrix and each of the pre-configured multiple target matrix information reference vectors (the target matrix information reference vectors can be obtained based on mining processing of the reference parameter distribution matrix) are calculated respectively, and then based on the vector similarities between each target matrix information reference vector and the target matrix information to-be-processed vector, the target air-conditioning performance labeling results corresponding to each target matrix information reference vector are fused (for example, the vector similarities can be used as corresponding weight values ​​to perform weighted sum calculation) to form the target air-conditioning performance prediction result corresponding to the target automobile air-conditioner.

[0076] Further supplementary explanation, in some embodiments, during the step of performing the key data extraction operation on the historical indicator parameter distribution matrix and outputting the key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix, the following steps may be performed instead:

[0077] Performing a numerical mapping operation on each parameter value of the historical air conditioning performance index parameter included in the historical index parameter distribution matrix (for example, the parameter value of each historical air conditioning performance index parameter included in the historical index parameter distribution matrix can be mapped to a value interval [0, 1]);

[0078] Replacing the parameter values ​​of the key historical air-conditioning performance index parameters in the historical index parameter distribution matrix after the numerical mapping operation with pre-configured key parameter values ​​(exemplarily, based on the parameter value mapping value interval [0, 1], the key parameter value may be 1), and replacing the parameter values ​​of the non-key historical air-conditioning performance index parameters in the historical index parameter distribution matrix after the numerical mapping operation with pre-configured non-key parameter values ​​(exemplarily, based on the parameter value mapping value interval [0, 1], the non-key parameter value may be 0), so as to form a key historical index parameter replacement distribution matrix corresponding to the historical index parameter distribution matrix, wherein the parameter values ​​corresponding to the key historical air-conditioning performance index parameters after the numerical mapping operation are not less than a pre-configured parameter comparison value (exemplarily, based on the parameter value mapping value interval [0, 1], the parameter comparison value may be 0.5), and the parameter values ​​corresponding to the non-key historical air-conditioning performance index parameters after the numerical mapping operation are less than the parameter comparison value;

[0079] A matrix fusion operation is performed on the key historical indicator parameter replacement distribution matrix and the historical indicator parameter distribution matrix (for example, based on the key parameter value being 1 and the non-key parameter value being 0, a matrix multiplication operation can be performed on the key historical indicator parameter replacement distribution matrix and the historical indicator parameter distribution matrix) to output the key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix.

[0080] Further supplementary explanation, in some embodiments, the step of analyzing and outputting the target matrix information to be processed vector corresponding to the historical indicator parameter distribution matrix based on the matrix overall information to be processed vector and the corresponding matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix can be replaced by performing the following steps during the execution of this step:

[0081] A first vector dimensionality reduction model (exemplarily, the first vector dimensionality reduction model belongs to the target data indicator prediction neural network) is used to perform a vector dimensionality reduction operation on the matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix (the vector dimensionality reduction operation can compress the dimension of the vector and make the feature representativeness stronger), and output the reduced dimensionality matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix. The first vector dimensionality reduction model performs a model parameter update operation on the to-be-updated vector dimensionality reduction model based on the corresponding key information mining error and / or information mining joint error (i.e., any one or the sum of the two errors) (as described below);

[0082] Perform vector aggregation operations (such as vector superposition or splicing operations) on the matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix to output the target matrix information to be processed vector corresponding to the historical indicator parameter distribution matrix.

[0083] Further supplementary explanation, in some embodiments, the data indicator prediction method for automobile air conditioners may further include a step of updating and forming the first matrix key information mining model. During the execution of this step, the following steps may be performed instead:

[0084] Extracting the exemplary data combination, loading the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix included in the exemplary data combination into the first matrix overall information mining model to complete the matrix overall information mining operation, and outputting the standard matrix overall information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the correlation matrix overall information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the non-correlation matrix overall information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix;

[0085] Analyze the key historical indicator parameter distribution matrices corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, respectively; load the key historical indicator parameter distribution matrices corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, respectively, into the matrix key information mining model to be updated (such as one constructed based on a deep residual network) to complete the matrix key information mining operation, and output the matrix key information to-be-processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, respectively;

[0086] According to the standard matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the corresponding standard target matrix information to be processed vector is analyzed and outputted; then, according to the correlation matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, the corresponding correlation target matrix information to be processed vector is analyzed and outputted; then, according to the non-correlation matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix, the corresponding non-correlation target matrix information to be processed vector is analyzed and outputted;

[0087] Analyze and output the corresponding key information mining error based on the matrix key information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the matrix key information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the matrix key information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix;

[0088] Analyze and output the corresponding information mining joint error based on the standard target matrix information to be processed vector, the relevant target matrix information to be processed vector, and the non-relevant target matrix information to be processed vector;

[0089] An update operation is performed on the model parameters of the matrix critical information mining model to be updated based on the critical information mining error and the information mining joint error (for example, the update operation can be first performed based on the critical information mining error, and then the update operation can be performed based on the information mining joint error) to form an updated matrix critical information mining model. The first matrix critical information mining model is obtained based on the updated matrix critical information mining model (for example, the updated matrix critical information mining model can be used as the first matrix critical information mining model, or the first matrix critical information mining model can be constructed based on the model parameters of the updated matrix critical information mining model).

[0090] Based on this, by fusing the matrix key information to be processed vector and the matrix overall information to be processed vector, the matrix key information mining model to be updated can use the target matrix information to be processed vector as the compatibility update data on the basis of key information learning, so that the obtained first matrix key information mining model can extract the matrix key information to be processed vector with higher accuracy, thereby improving the accuracy of the target matrix information to be processed vector.

[0091] Further supplementary explanation, in some embodiments, in the process of executing the steps of analyzing and outputting the corresponding standard target matrix information to be processed vector based on the standard matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, then analyzing and outputting the corresponding relevant target matrix information to be processed vector based on the relevant matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, and then analyzing and outputting the corresponding non-relevant target matrix information to be processed vector based on the non-relevant matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix, the following steps may be performed instead:

[0092] Loading the matrix key information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the matrix key information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the matrix key information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix into the vector dimensionality reduction model to be updated to complete the vector dimensionality reduction operation, and outputting the reduced dimensionality matrix key information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the reduced dimensionality matrix key information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the reduced dimensionality matrix key information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix;

[0093] According to the standard matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix (exemplarily, the exemplary standard indicator parameter distribution matrix and the corresponding dimensionality reduction matrix key information to be processed vector can be vector spliced), the corresponding standard target matrix information to be processed vector is analyzed and output, and then, according to the relevant matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix (exemplarily, the exemplary first indicator parameter distribution matrix and the corresponding dimensionality reduction matrix key information to be processed vector can be vector spliced), the corresponding relevant target matrix information to be processed vector is analyzed and output, and, then, according to the non-relevant matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix (exemplarily, the exemplary second indicator parameter distribution matrix and the corresponding dimensionality reduction matrix key information to be processed vector can be vector spliced), the corresponding non-relevant target matrix information to be processed vector is analyzed and output.

[0094] Further supplementary explanation, in some embodiments, in the process of executing the step of analyzing and outputting the corresponding critical information mining errors based on the matrix critical information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the matrix critical information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the matrix critical information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix, the following steps may be performed instead:

[0095] Calculating the vector distance between the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix and the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix to output the corresponding correlated key vector distance, and then calculating the vector distance between the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix and the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix to output the corresponding non-correlated key vector distance;

[0096] Analyzing and outputting a corresponding joint key information mining error according to a degree of difference between the relevant key vector distance and the non-relevant key vector distance;

[0097] Parameter mapping operations are performed on the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix (for example, the vector parameters in the matrix key information to be processed vector can be mapped to the interval [-1,1]) to output the corresponding mapping indicator parameters respectively, and then according to the output of each mapping indicator parameter, the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix, the corresponding mapping key information mining error is analyzed and output (for example, the vector parameters in the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix can be mapped to the interval [-1,1]). The square of the difference between the matrix key information to be processed and the corresponding mapping index parameter can be calculated, and the square calculation results can be summed to obtain the corresponding standard mapping key information mining error; the square of the difference between the vector parameters in the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix and the corresponding mapping index parameter can be calculated, and the square calculation results can be summed to obtain the corresponding first mapping key information mining error; the square of the difference between the vector parameters in the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix and the corresponding mapping index parameter can be calculated, and the square calculation results can be summed to obtain the corresponding second mapping key information mining error; then, the standard mapping key information mining error, the first mapping key information mining error and the second mapping key information mining error are averaged, and the result of the average calculation is used as the mapping key information mining error);

[0098] According to the joint key information mining error and the mapped key information mining error, the key information mining error corresponding to the fusion output (such as weighted mean calculation, etc.) is obtained.

[0099] Further supplementary explanation, in some embodiments, in the process of performing the step of analyzing and outputting the corresponding information mining joint error based on the standard target matrix information to be processed vector, the relevant target matrix information to be processed vector, and the non-relevant target matrix information to be processed vector, the following steps may be performed instead:

[0100] Calculating the vector distance between the standard target matrix information to be processed vector and the relevant target matrix information to be processed vector to output the corresponding relevant vector distance, and then calculating the vector distance between the standard target matrix information to be processed vector and the non-relevant target matrix information to be processed vector to output the corresponding non-relevant vector distance;

[0101] Based on the vector distance between the to-be-processed vector of the standard matrix overall information and the to-be-processed vector of the correlation matrix overall information, and in combination with the vector distance between the to-be-processed vector of the standard matrix overall information and the to-be-processed vector of the non-correlation matrix overall information, a corresponding vector distance difference reference value is calculated and output (for example, a difference calculation can be performed on the two preceding vector distances, and then a vector distance difference reference value having a positive correlation is determined based on the result of the difference calculation);

[0102] Based on the degree of distance difference between the relevant vector distance and the non-relevant vector distance, and combined with the vector distance difference reference value, the corresponding information mining joint error is calculated and output (for example, the matrix difference degree and the vector distance difference reference value can be summed or averaged to obtain the information mining joint error).

[0103] Further supplementary explanation, in some embodiments, during the process of executing step S130, the following steps may be performed instead:

[0104] Extracting multiple target air conditioning performance index parameters of the target automobile air conditioner at the current time, performing weighted fusion processing on the multiple target air conditioning performance index parameters (the specific weighting coefficients can be configured according to actual application requirements and are not specifically limited here), and outputting a calculation result of the actual air conditioning performance corresponding to the target automobile air conditioner;

[0105] A weighted summation is performed on the actual air-conditioning performance calculation result and the target air-conditioning performance prediction result (the specific weighting coefficient can be configured according to actual application requirements, for example, the weight corresponding to the actual air-conditioning performance calculation result is less than the weight corresponding to the target air-conditioning performance prediction result), and a target air-conditioning performance fusion result corresponding to the target automobile air-conditioning is obtained.

[0106] Combine Figure 3 The embodiment of the present invention further provides a data indicator prediction device for automobile air conditioners, which can be applied to the above-mentioned data indicator prediction system for automobile air conditioners. The data indicator prediction device for automobile air conditioners may include:

[0107] a performance index parameter extraction module for extracting, for each of a plurality of historical time values ​​included in a target time period, a plurality of historical air conditioning performance index parameters of the target automobile air conditioner at the historical time value;

[0108] a data indicator prediction module, configured to use the updated target data indicator prediction neural network to analyze and process a plurality of historical air conditioning performance indicator parameters corresponding to the plurality of historical time values, so as to output a target air conditioning performance prediction result corresponding to the target automobile air conditioner, wherein the target air conditioning performance prediction result is used to reflect the degree of excellence of the comprehensive performance of the target automobile air conditioner;

[0109] The air conditioning performance fusion module is used to extract multiple target air conditioning performance index parameters of the target automobile air conditioning at the current time, and obtain the target air conditioning performance fusion result corresponding to the target automobile air conditioning based on the multiple target air conditioning performance index parameters and the target air conditioning performance prediction result.

[0110] In summary, the present invention provides a data indicator prediction method and system for an automotive air conditioner. For each of multiple historical time values ​​included in a target time period, multiple historical air conditioning performance indicator parameters of the target automotive air conditioner at that historical time value are extracted. An updated target data indicator prediction neural network is used to analyze and process the multiple historical air conditioning performance indicator parameters corresponding to the multiple historical time values ​​to output a target air conditioning performance prediction result corresponding to the target automotive air conditioner. Multiple target air conditioning performance indicator parameters of the target automotive air conditioner at the current time are extracted, and a corresponding target air conditioning performance fusion result is obtained based on the multiple target air conditioning performance indicator parameters and the target air conditioning performance prediction result. Based on the foregoing, the target data indicator prediction neural network can be used to first comprehensively analyze a large number of historical air conditioning performance indicator parameters to output a corresponding target air conditioning performance prediction result. The multiple current target air conditioning performance indicator parameters are then fused with the target air conditioning performance prediction result to obtain a corresponding target air conditioning performance fusion result. This provides a more comprehensive basis for analyzing (predicting) the data indicators, thereby improving the reliability of the automotive air conditioner data indicator prediction to a certain extent.

[0111] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A data index prediction method for automobile air conditioners, characterized in that: include: For each of the multiple historical time values ​​included in the target time period, extracting multiple historical air-conditioning performance index parameters of the target automobile air-conditioner at the historical time value; Using the updated target data indicator prediction neural network, a plurality of historical air conditioning performance indicator parameters corresponding to the plurality of historical time values ​​are analyzed and processed to output a target air conditioning performance prediction result corresponding to the target automobile air conditioner, the target air conditioning performance prediction result being used to reflect the degree of excellence of the comprehensive performance of the target automobile air conditioner; Extracting multiple target air conditioning performance index parameters of the target automobile air conditioner at the current time, and obtaining a target air conditioning performance fusion result corresponding to the target automobile air conditioner based on the multiple target air conditioning performance index parameters and the target air conditioning performance prediction result; The step of using the updated target data indicator prediction neural network to analyze and process the multiple historical air conditioning performance indicator parameters corresponding to the multiple historical time values ​​to output a target air conditioning performance prediction result corresponding to the target automobile air conditioner includes: Matrixing the multiple historical air conditioning performance index parameters corresponding to the multiple historical time values ​​to form a corresponding historical index parameter distribution matrix, wherein a row of the historical air conditioning performance index parameters is a value of a historical air conditioning performance index parameter at the multiple historical time values, and a column of the historical air conditioning performance index parameters is a value of multiple historical air conditioning performance index parameters at one historical time value; Using a target data indicator prediction neural network, the historical indicator parameter distribution matrix is ​​analyzed and processed to output a target air-conditioning performance prediction result corresponding to the target automobile air-conditioning; The step of using the target data indicator prediction neural network to analyze and process the historical indicator parameter distribution matrix to output a target air conditioning performance prediction result corresponding to the target automobile air conditioner includes: Using the first matrix overall information mining model included in the updated target data indicator prediction neural network to perform a matrix overall information mining operation on the historical indicator parameter distribution matrix to output a matrix overall information to-be-processed vector corresponding to the historical indicator parameter distribution matrix; Performing a key data extraction operation on the historical indicator parameter distribution matrix, and outputting a key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix; Using the first matrix key information mining model included in the target data indicator prediction neural network to perform a matrix key information mining operation on the key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix, so as to output a matrix key information to-be-processed vector corresponding to the historical indicator parameter distribution matrix; Analyze and output the target matrix information to be processed vector corresponding to the historical indicator parameter distribution matrix based on the matrix overall information to be processed vector and the corresponding matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix; respectively calculating the vector similarity between the target matrix information to-be-processed vector corresponding to the historical indicator parameter distribution matrix and each of the pre-configured multiple target matrix information reference vectors, and then, based on the vector similarity between each target matrix information reference vector and the target matrix information to-be-processed vector, fusing the target air-conditioning performance labeling result corresponding to each target matrix information reference vector to form a target air-conditioning performance prediction result corresponding to the target automobile air-conditioning; The first matrix key information mining model is formed by updating model parameters of the matrix key information mining model to be updated based on the first matrix overall information mining model and the extracted exemplary data combination, wherein the exemplary data combination includes an exemplary standard indicator parameter distribution matrix, an exemplary first indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix, and an exemplary second indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix; In the process of updating the model parameters of the matrix critical information mining model to be updated, the first matrix overall information mining model is used to perform a matrix overall information mining operation on the matrix overall information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix, and the matrix critical information mining model to be updated is used to perform a matrix critical information mining operation on the matrix critical information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix. The first matrix critical information mining model updates the model parameters of the matrix critical information mining model to be updated based on the key information mining error and information mining joint error output by the analysis. The key information mining error is output according to the analysis of the matrix critical information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix respectively. The information mining joint error is output according to the analysis of the matrix overall information to be processed vectors and the matrix critical information to be processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix respectively. The matrix similarity between the exemplary standard indicator parameter distribution matrix and the exemplary first indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix is ​​greater than the matrix similarity reference value, and the matrix similarity between the exemplary standard indicator parameter distribution matrix and the exemplary second indicator parameter distribution matrix corresponding to the exemplary standard indicator parameter distribution matrix is ​​less than the matrix similarity reference value.

2. The data indicator prediction method for automobile air conditioner according to claim 1, characterized in that: The step of extracting, for each of the multiple historical time values ​​included in the target time period, multiple historical air conditioning performance index parameters of the target automobile air conditioner at the historical time value includes: Obtaining the most recently analyzed historical target air conditioning performance fusion result, and then updating the pre-configured reference time period duration based on the historical target air conditioning performance fusion result to output a corresponding target time period duration; At the current time, a corresponding target time period is determined based on the target time period duration, wherein the target time period duration is equal to the target time period duration, and the target time period end point is the current time; For each historical time value included in the target time period, a plurality of historical air-conditioning performance index parameters of the target automobile air-conditioner at the historical time value are extracted. The target time period includes a plurality of historical time values.

3. The data indicator prediction method for automobile air conditioner according to claim 1, characterized in that: The step of performing a key data extraction operation on the historical indicator parameter distribution matrix and outputting a key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix includes: Performing a numerical mapping operation on each parameter value of the historical air-conditioning performance index parameter included in the historical index parameter distribution matrix; Replacing parameter values ​​of key historical air-conditioning performance index parameters in the historical index parameter distribution matrix after the numerical mapping operation with pre-configured key parameter values, and replacing parameter values ​​of non-key historical air-conditioning performance index parameters in the historical index parameter distribution matrix after the numerical mapping operation with pre-configured non-key parameter values, so as to form a key historical index parameter replacement distribution matrix corresponding to the historical index parameter distribution matrix, wherein the parameter values ​​corresponding to the key historical air-conditioning performance index parameters after the numerical mapping operation are not less than the pre-configured parameter comparison values, and the parameter values ​​corresponding to the non-key historical air-conditioning performance index parameters after the numerical mapping operation are less than the parameter comparison value; A matrix fusion operation is performed on the key historical indicator parameter replacement distribution matrix and the historical indicator parameter distribution matrix to output a key historical indicator parameter distribution matrix corresponding to the historical indicator parameter distribution matrix.

4. The data indicator prediction method for automobile air conditioner according to claim 1, characterized in that: The data indicator prediction method further includes: Extracting the exemplary data combination, loading the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix included in the exemplary data combination into the first matrix overall information mining model to complete the matrix overall information mining operation, and outputting the standard matrix overall information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the correlation matrix overall information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the non-correlation matrix overall information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix; Analyzing the key historical indicator parameter distribution matrices corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, respectively, loading the key historical indicator parameter distribution matrices corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, respectively, into the matrix key information mining model to be updated to complete the matrix key information mining operation, and outputting the matrix key information to-be-processed vectors corresponding to the exemplary standard indicator parameter distribution matrix, the exemplary first indicator parameter distribution matrix, and the exemplary second indicator parameter distribution matrix, respectively; According to the standard matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the corresponding standard target matrix information to be processed vector is analyzed and outputted; then, according to the correlation matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, the corresponding correlation target matrix information to be processed vector is analyzed and outputted; then, according to the non-correlation matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix, the corresponding non-correlation target matrix information to be processed vector is analyzed and outputted; Analyze and output the corresponding key information mining error based on the matrix key information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the matrix key information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the matrix key information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix; Analyze and output the corresponding information mining joint error based on the standard target matrix information to be processed vector, the relevant target matrix information to be processed vector, and the non-relevant target matrix information to be processed vector; The model parameters of the matrix critical information mining model to be updated are updated according to the critical information mining error and the information mining joint error to form an updated matrix critical information mining model, and the first matrix critical information mining model is obtained according to the updated matrix critical information mining model.

5. The data index prediction method for automobile air conditioner according to claim 4, characterized in that: The steps of analyzing and outputting the corresponding standard target matrix information to be processed vector based on the standard matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, then analyzing and outputting the corresponding relevant target matrix information to be processed vector based on the relevant matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, and then analyzing and outputting the corresponding non-relevant target matrix information to be processed vector based on the non-relevant matrix overall information to be processed vector and the matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix, include: Loading the matrix key information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the matrix key information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the matrix key information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix into the vector dimensionality reduction model to be updated to complete the vector dimensionality reduction operation, and outputting the reduced dimensionality matrix key information to-be-processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the reduced dimensionality matrix key information to-be-processed vector corresponding to the exemplary first indicator parameter distribution matrix, and the reduced dimensionality matrix key information to-be-processed vector corresponding to the exemplary second indicator parameter distribution matrix; Based on the standard matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the exemplary standard indicator parameter distribution matrix, the corresponding standard target matrix information to be processed vector is analyzed and output; then, based on the relevant matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the exemplary first indicator parameter distribution matrix, the corresponding relevant target matrix information to be processed vector is analyzed and output; and, based on the non-relevant matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the exemplary second indicator parameter distribution matrix, the corresponding non-relevant target matrix information to be processed vector is analyzed and output.

6. The data indicator prediction method for automobile air conditioner according to claim 5, characterized in that: The step of analyzing and outputting a target matrix information to be processed vector corresponding to the historical indicator parameter distribution matrix based on the matrix overall information to be processed vector and the corresponding matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix includes: A first vector dimensionality reduction model is used to perform a vector dimensionality reduction operation on the matrix key information to-be-processed vector corresponding to the historical indicator parameter distribution matrix, and output a dimensionality reduction matrix key information to-be-processed vector corresponding to the historical indicator parameter distribution matrix, wherein the first vector dimensionality reduction model is formed by performing a model parameter update operation on the vector dimensionality reduction model to be updated based on the key information mining error and / or the information mining joint error; A vector aggregation operation is performed on the matrix overall information to be processed vector and the dimensionality reduction matrix key information to be processed vector corresponding to the historical indicator parameter distribution matrix to output the target matrix information to be processed vector corresponding to the historical indicator parameter distribution matrix.

7. The data indicator prediction method for automobile air conditioner according to claim 6, characterized in that: The step of extracting a plurality of target air conditioning performance index parameters of the target automobile air conditioner at the current time, and obtaining a target air conditioning performance fusion result corresponding to the target automobile air conditioner based on the plurality of target air conditioning performance index parameters and the target air conditioning performance prediction result, comprises: Extracting multiple target air conditioning performance index parameters of the target automobile air conditioner at the current time, performing weighted fusion processing on the multiple target air conditioning performance index parameters to output a calculation result of the actual air conditioning performance corresponding to the target automobile air conditioner; A weighted sum calculation is performed on the actual air-conditioning performance calculation result and the target air-conditioning performance prediction result to obtain a target air-conditioning performance fusion result corresponding to the target automobile air-conditioner.

8. A data index prediction system for automobile air conditioners, characterized in that: It includes a processor and a memory, the memory is used to store a computer program, and the processor is used to execute the computer program to implement the data indicator prediction method described in any one of claims 1-7.

Images