Automobile air conditioning energy consumption prediction method and system

Abstract

The present invention provides a method and system for predicting automobile air conditioning energy consumption, relating to the field of data processing technology. In this method, target automobile air conditioning environment data corresponding to a target automobile air conditioning unit is collected, the target automobile air conditioning environment data being used to reflect the environment of the space where the target automobile air conditioning unit is located. The target automobile air conditioning environment data is then loaded into an optimized target air conditioning energy consumption prediction neural network, which uses the target air conditioning energy consumption prediction neural network to perform energy consumption prediction processing on the target automobile air conditioning environment data and output a corresponding target energy consumption prediction result. The actual air conditioning energy consumption value of the target automobile air conditioning unit within a recent time period is extracted, and the actual air conditioning energy consumption value and the target energy consumption prediction result are fused to output a target energy consumption fusion result corresponding to the target automobile air conditioning unit. Based on the above, the reliability of automobile air conditioning energy consumption prediction can be improved to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for predicting energy consumption of an automobile air conditioner. Background Art

[0002] Predicting automotive air conditioning energy consumption is a critical component of many automotive applications, and therefore requires reliable predictions. However, existing technologies typically use historical energy consumption data to predict current energy consumption. This limited data, coupled with insufficient predictive analysis, can lead to unreliable predictions. Summary of the Invention

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

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

[0005] A method for predicting energy consumption of an automobile air conditioner, comprising:

[0006] collecting target automobile air-conditioning environment data corresponding to the target automobile air-conditioning, wherein the target automobile air-conditioning environment data is used to reflect the environment of the space where the target automobile air-conditioning is located;

[0007] The target automobile air-conditioning environment data is loaded into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, so as to perform energy consumption prediction processing on the target automobile air-conditioning environment data using the target air-conditioning energy consumption prediction neural network and output a corresponding target energy consumption prediction result;

[0008] The actual air conditioning energy consumption value of the target automobile air conditioner in a recent time period is extracted, and the actual air conditioning energy consumption value and the target energy consumption prediction result are fused to output a target energy consumption fusion result corresponding to the target automobile air conditioner.

[0009] In some preferred embodiments, in the above-mentioned automobile air-conditioning energy consumption prediction method, the step of collecting target automobile air-conditioning environment data corresponding to the target automobile air-conditioning includes:

[0010] Extracting and processing the navigation data of the target vehicle corresponding to the target vehicle air conditioner to output target navigation data corresponding to the target vehicle, and then performing data analysis and processing on the target navigation data to determine a target navigation path corresponding to the target navigation data;

[0011] Environmental data collection and processing are performed on each path area on the target navigation path to form target automobile air-conditioning environmental data corresponding to the target automobile air-conditioning.

[0012] In some preferred embodiments, in the above-mentioned automobile air conditioning energy consumption prediction method, the step of loading the target automobile air conditioning environment data into a target air conditioning energy consumption prediction neural network that has undergone an optimized operation, performing energy consumption prediction processing on the target automobile air conditioning environment data using the target air conditioning energy consumption prediction neural network, and outputting a corresponding target energy consumption prediction result includes:

[0013] The target automobile air-conditioning environment data is loaded into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, so as to use the target air-conditioning energy consumption prediction neural network to perform energy consumption prediction processing on the target automobile air-conditioning environment data, output first matching probabilities corresponding to the target automobile air-conditioning environment data in a plurality of initial energy consumption label information, and output updated second matching probabilities corresponding to the target automobile air-conditioning environment data in at least one extended energy consumption label information, wherein the plurality of initial energy consumption label information and the at least one extended energy consumption label information are respectively configured based on exemplary automobile air-conditioning environment data of different time periods, and the data formation time of the exemplary automobile air-conditioning environment data corresponding to the initial energy consumption label information is earlier than the data formation time of the exemplary automobile air-conditioning environment data corresponding to the extended energy consumption label information, and each of the initial energy consumption label information and each of the initial energy consumption label information are different;

[0014] Based on the first matching probabilities corresponding to the multiple initial energy consumption label information and the second matching probabilities corresponding to the at least one extended energy consumption label information, the matching energy consumption label information corresponding to the target automobile air-conditioning environment data is matched from the multiple initial energy consumption label information and the at least one extended energy consumption label information, and then the energy consumption value reflected by the matching energy consumption label information is marked as the corresponding target energy consumption prediction result.

[0015] In some preferred embodiments, in the above-mentioned automobile air-conditioning energy consumption prediction method, the step of loading the target automobile air-conditioning environment data into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, performing energy consumption prediction processing on the target automobile air-conditioning environment data using the target air-conditioning energy consumption prediction neural network, outputting first matching probabilities corresponding to the target automobile air-conditioning environment data in a plurality of initial energy consumption label information, and outputting updated second matching probabilities corresponding to the target automobile air-conditioning environment data in at least one extended energy consumption label information, includes:

[0016] The target vehicle air-conditioning environment data is loaded into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, and a data mining operation is performed on the target vehicle air-conditioning environment data using the target air-conditioning energy consumption prediction neural network to output a corresponding first environment data mining result, and then, based on the first environment data mining result, a first matching probability corresponding to the target vehicle air-conditioning environment data in the at least one extended energy consumption label information and the plurality of initial energy consumption label information is analyzed and output;

[0017] The target air conditioner energy consumption prediction neural network is used, and based on the reference probability update parameter, the first matching probability corresponding to the at least one extended energy consumption label information is updated to output the second matching probability corresponding to the at least one extended energy consumption label information.

[0018] In some preferred embodiments, in the above-mentioned automobile air conditioning energy consumption prediction method, the automobile air conditioning energy consumption prediction method further includes the step of optimizing the target air conditioning energy consumption prediction neural network, which step includes:

[0019] Using the extracted first example environment data set and the second example environment data set, a network optimization operation is performed on the initial air conditioning energy consumption prediction neural network, so as to output an intermediate air conditioning energy consumption prediction neural network when a learning cost value formed during the network optimization process is less than a pre-configured learning cost value threshold, wherein the first example environment data set corresponds to at least one extended energy consumption label information, and the second example environment data set corresponds to multiple initial energy consumption label information;

[0020] Extracting multiple updated exemplary automobile air-conditioning environment data, then loading each of the updated exemplary automobile air-conditioning environment data into the intermediate air-conditioning energy consumption prediction neural network, using the intermediate air-conditioning energy consumption prediction neural network to perform energy consumption prediction processing, outputting a matching probability of the first extended energy consumption label information corresponding to the corresponding updated exemplary automobile air-conditioning environment data in the at least one extended energy consumption label information, and outputting a matching probability of the first initial energy consumption label information corresponding to the corresponding updated exemplary automobile air-conditioning environment data in the multiple initial energy consumption label information, each of the updated exemplary automobile air-conditioning environment data corresponding to one of the extended energy consumption label information or one of the initial energy consumption label information;

[0021] Based on the predicted matching probability of the first extended energy consumption label information and the first initial energy consumption label information, the corresponding reference probability update parameters are analyzed and output, and then the reference probability update parameters are used to update the output of the intermediate air-conditioning energy consumption prediction neural network to form the corresponding target air-conditioning energy consumption prediction neural network.

[0022] In some preferred embodiments, in the above-mentioned automobile air conditioning energy consumption prediction method, the steps of analyzing and outputting corresponding reference probability update parameters based on the predicted output matching probability of the first extended energy consumption label information and the first initial energy consumption label information, and then using the reference probability update parameters to update the output of the intermediate air conditioning energy consumption prediction neural network to form a corresponding target air conditioning energy consumption prediction neural network include:

[0023] Calculate and output an extended energy consumption label representative value corresponding to the at least one extended energy consumption label information based on the predicted matching probability of the first extended energy consumption label information;

[0024] Calculate and output initial energy consumption label representative values ​​corresponding to the plurality of initial energy consumption label information based on the predicted matching probability of the first initial energy consumption label information;

[0025] Based on the extended energy consumption label representative value and the initial energy consumption label representative value, calculating and outputting a corresponding reference probability update parameter;

[0026] The reference probability update parameter is configured in the output subnetwork of the intermediate air conditioning energy consumption prediction neural network to form a corresponding target air conditioning energy consumption prediction neural network, and the reference probability update parameter is used to update the matching probability of the extended energy consumption label information corresponding to the at least one extended energy consumption label information.

[0027] In some preferred embodiments, in the above-mentioned automobile air conditioning energy consumption prediction method, the step of performing a network optimization operation on the initial air conditioning energy consumption prediction neural network using the extracted first example environment data set and the second example environment data set, so as to output an intermediate air conditioning energy consumption prediction neural network when a learning cost value formed during the network optimization process is less than a pre-configured learning cost value threshold, includes:

[0028] Loading each exemplary extended automobile air-conditioning environment data in the first exemplary environment data set into an initial air-conditioning energy consumption prediction neural network, performing analysis and prediction processing using the initial air-conditioning energy consumption prediction neural network, outputting a matching probability of the corresponding exemplary extended automobile air-conditioning environment data with the second extended energy consumption label information corresponding to the at least one extended energy consumption label information, and outputting a matching probability of the corresponding exemplary extended automobile air-conditioning environment data with the second initial energy consumption label information corresponding to the multiple initial energy consumption label information;

[0029] Loading each exemplary initial automobile air-conditioning environment data in the second exemplary environment data set into the initial air-conditioning energy consumption prediction neural network, performing analysis and prediction processing using the initial air-conditioning energy consumption prediction neural network, outputting a matching probability of the corresponding exemplary initial automobile air-conditioning environment data with the third extended energy consumption label information corresponding to the at least one extended energy consumption label information, and outputting a matching probability of the corresponding exemplary initial automobile air-conditioning environment data with the third initial energy consumption label information corresponding to the multiple initial energy consumption label information;

[0030] Based on the analysis output of the second extended energy consumption label information matching probability, the second initial energy consumption label information matching probability, the third extended energy consumption label information matching probability and the third initial energy consumption label information matching probability, the corresponding network optimization learning cost value is calculated and output, and then the network optimization operation is performed on the initial air-conditioning energy consumption prediction neural network based on the network optimization learning cost value, and when the learning cost value formed during the network optimization process is less than the pre-configured learning cost value threshold, the intermediate air-conditioning energy consumption prediction neural network corresponding to the initial air-conditioning energy consumption prediction neural network is output.

[0031] In some preferred embodiments, in the above-mentioned automobile air conditioner energy consumption prediction method, the step of calculating and outputting the corresponding network optimization learning cost value based on the analyzed second extended energy consumption label information matching probability, the second initial energy consumption label information matching probability, the third extended energy consumption label information matching probability, and the third initial energy consumption label information matching probability includes:

[0032] Calculate and output a first-category learning cost based on the second extended energy consumption label information matching probability, the second initial energy consumption label information matching probability, the third extended energy consumption label information matching probability, and the third initial energy consumption label information matching probability outputted from the analysis;

[0033] Calculate and output a second type of learning cost value based on the second initial energy consumption label information matching probability and the third initial energy consumption label information matching probability outputted from the analysis;

[0034] The first type of learning cost value and the second type of learning cost value are integrated to output a corresponding network optimization learning cost value.

[0035] In some preferred embodiments, in the above-mentioned automobile air conditioner energy consumption prediction method, the step of extracting the actual air conditioning energy consumption value of the target automobile air conditioner in a recent time period, and fusing the actual air conditioning energy consumption value with the target energy consumption prediction result to output the target energy consumption fusion result corresponding to the target automobile air conditioner includes:

[0036] Extracting, based on a pre-configured effective time length, an actual air conditioning energy consumption value of the target automobile air conditioner in a most recent effective time period, the actual air conditioning energy consumption value being used to reflect the unit energy consumption of the target automobile air conditioner in the most recent time period;

[0037] The actual air-conditioning energy consumption value and the unit predicted energy consumption reflected by the target energy consumption prediction result are weighted and calculated to output a target energy consumption fusion result corresponding to the target automobile air-conditioning.

[0038] An embodiment of the present invention further provides an automobile air conditioning energy consumption prediction system, comprising 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 automobile air conditioning energy consumption prediction method.

[0039] The embodiment of the present invention provides a method and system for predicting the energy consumption of an automobile air conditioner. The method collects target automobile air conditioner environmental data corresponding to a target automobile air conditioner; loads the target automobile air conditioner environmental data into a target air conditioner energy consumption prediction neural network that has undergone an optimized operation, and uses the target air conditioner energy consumption prediction neural network to perform energy consumption prediction processing on the target automobile air conditioner environmental data, and outputs a corresponding target energy consumption prediction result; extracts the actual air conditioner energy consumption value of the target automobile air conditioner in a recent time period, and fuses the actual air conditioner energy consumption value with the target energy consumption prediction result to output a target energy consumption fusion result corresponding to the target automobile air conditioner. Based on this, when predicting and analyzing air conditioner energy consumption, not only the historical energy consumption of the automobile air conditioner itself is considered, but also the results of predictive analysis of environmental data based on the neural network are considered, realizing the fusion of multi-dimensional data, making the basis for predictive analysis sufficient, thereby improving the reliability of automobile air conditioner energy consumption prediction to a certain extent.

[0040] 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

[0041] Figure 1 This is a structural block diagram of the automobile air conditioning energy consumption prediction system provided by an embodiment of the present invention.

[0042] Figure 2 A flowchart illustrating the steps of a method for predicting energy consumption of an automobile air conditioner provided in an embodiment of the present invention.

[0043] Figure 3 A schematic diagram of the modules included in the automobile air-conditioning energy consumption prediction device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] 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.

[0045] like Figure 1 As shown, an embodiment of the present invention provides an automobile air-conditioning energy consumption prediction system, wherein the automobile air-conditioning energy consumption prediction system may include a memory and a processor.

[0046] For example, in some more detailed embodiments, the memory and processor are directly or indirectly electrically connected to enable data transmission or exchange. For example, the electrical connection may be achieved via one or more communication buses or signal lines. The memory may store at least one software function module (computer program) in the form of software or firmware. The processor may be configured to execute the executable computer program stored in the memory, thereby implementing the automotive air conditioner energy consumption prediction method provided in embodiments of the present invention.

[0047] For example, in some more detailed 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.

[0048] For example, in some more detailed implementations, the automobile air-conditioning energy consumption prediction system may be a server with data processing capabilities.

[0049] Combine Figure 2 The present invention also provides an automotive air conditioning energy consumption prediction method, which can be applied to the automotive air conditioning energy consumption prediction system. The steps defined in the process flow related to the automotive air conditioning energy consumption prediction method can be implemented by the automotive air conditioning energy consumption prediction system.

[0050] The following will Figure 2 The specific process shown is explained in detail.

[0051] Step S110 , collecting target automobile air-conditioning environment data corresponding to the target automobile air-conditioning.

[0052] In an embodiment of the present invention, the automotive air conditioner energy consumption prediction system can collect target automotive air conditioner environmental data corresponding to a target automotive air conditioner. The target automotive air conditioner environmental data reflects the environment of the space where the target automotive air conditioner is located (e.g., ambient temperature, ambient humidity, ambient wind speed, ambient light intensity, ambient rainfall, etc.).

[0053] Step S120, loading the target automobile air-conditioning environment data into the target air-conditioning energy consumption prediction neural network that has undergone optimization operation, so as to use the target air-conditioning energy consumption prediction neural network to perform energy consumption prediction processing on the target automobile air-conditioning environment data and output the corresponding target energy consumption prediction result.

[0054] In an embodiment of the present invention, the automobile air-conditioning energy consumption prediction system can load the target automobile air-conditioning environment data into a target air-conditioning energy consumption prediction neural network that has undergone optimized operations, so as to use the target air-conditioning energy consumption prediction neural network to perform energy consumption prediction processing on the target automobile air-conditioning environment data and output the corresponding target energy consumption prediction result.

[0055] Step S130 , extracting the actual air conditioning energy consumption value of the target automobile air conditioner in a recent time period, and fusing the actual air conditioning energy consumption value with the target energy consumption prediction result to output a target energy consumption fusion result corresponding to the target automobile air conditioner.

[0056] In an embodiment of the present invention, the automobile air-conditioning energy consumption prediction system can extract the actual air-conditioning energy consumption value of the target automobile air-conditioning in a recent time period, and fuse the actual air-conditioning energy consumption value and the target energy consumption prediction result to output the target energy consumption fusion result corresponding to the target automobile air-conditioning.

[0057] Based on this, when predicting and analyzing air conditioning energy consumption, not only the historical energy consumption of the automobile air conditioner itself will be considered, but also the results of predictive analysis of environmental data based on neural networks, realizing the fusion of multi-dimensional data, making the basis for predictive analysis sufficient, thereby improving the reliability of automobile air conditioning energy consumption prediction to a certain extent, and improving the deficiencies in existing technologies.

[0058] For example, in some more detailed implementations, in order to implement step S110, the following achievable sub-steps may be further performed:

[0059] Extracting and processing the navigation data of the target vehicle corresponding to the target vehicle air conditioner to output target navigation data corresponding to the target vehicle, and then performing data analysis and processing on the target navigation data to determine a target navigation path corresponding to the target navigation data;

[0060] Environmental data is collected and processed for each path area on the target navigation path (at least one path area can be divided according to road sections, that is, one road section corresponds to one path area) to form target automobile air-conditioning environmental data corresponding to the target automobile air-conditioning.

[0061] For example, in some more detailed implementations, in order to implement the step of collecting and processing environmental data for each path area on the target navigation path to form target automobile air-conditioning environmental data corresponding to the target automobile air-conditioning, the following achievable sub-steps may be further performed:

[0062] extracting each original path area on the target navigation path to obtain at least one original path area corresponding to the target navigation path, and, when the number of the at least one original path area is less than a preconfigured reference value, using each original path area as a path area, or, when the number of the at least one original path area is greater than or equal to the reference value, filtering the original path areas to obtain path areas;

[0063] Based on the environmental data of each path area obtained in the previous step, target automobile air-conditioning environmental data corresponding to the target automobile air-conditioning is formed by combining.

[0064] For example, in some more detailed implementations, in order to implement the step of screening the original path area to obtain the path area, the following achievable sub-steps may be further performed:

[0065] For each of the original route areas, extract historical unit energy consumption values ​​of automobile air conditioners of historical vehicles that have historically passed through the original route area, then perform an average calculation on the unit energy consumption values ​​to output the average unit energy consumption corresponding to the original route area, and extract the path area representative three-dimensional coordinates of the original route area (such as the three-dimensional coordinates of the area center point of the original route area, or the three-dimensional coordinates of the area point corresponding to the maximum value of the historical travel time of each area point in the original route area). Then, based on the average unit energy consumption and the path area representative three-dimensional coordinates, construct the area four-dimensional coordinates corresponding to the original route area;

[0066] Based on the regional four-dimensional coordinates corresponding to each of the original path regions, a regional coordinate distribution network corresponding to the original path region is constructed and outputted, and the regional coordinate distribution network is then segmented to form a set of regional coordinate distribution sub-networks corresponding to the regional coordinate distribution network, wherein the set of regional coordinate distribution sub-networks includes at least one regional coordinate distribution sub-network, and each of the regional coordinate distribution sub-networks includes at least one regional four-dimensional coordinate;

[0067] The step of performing segmentation processing on the regional coordinate distribution network to form a set of regional coordinate distribution sub-networks corresponding to the regional coordinate distribution network is performed multiple times to form a plurality of the regional coordinate distribution sub-network sets, and then for each of the regional coordinate distribution sub-network sets, the number of regional four-dimensional coordinates included in each regional coordinate distribution sub-network included in the regional coordinate distribution sub-network set is counted to output the number of network coordinates corresponding to each regional coordinate distribution sub-network, and the discreteness of the number of network coordinates corresponding to each regional coordinate distribution sub-network is calculated to output the discreteness of the number of coordinates corresponding to the regional coordinate distribution sub-network set;

[0068] For each of the regional coordinate distribution subnetwork sets, respectively calculate the set dispersion corresponding to each of the regional coordinate distribution subnetworks included in the regional coordinate distribution subnetwork set (for example, for each of the regional coordinate distribution subnetworks, the network center coordinates of the regional coordinate distribution subnetwork can be first determined, and then the coordinate distances between each regional four-dimensional coordinate in the regional coordinate distribution subnetwork and the network center coordinates can be respectively calculated, and the mean of the coordinate distances is calculated to output the set dispersion corresponding to the regional coordinate distribution subnetwork), and perform a dispersion calculation on the set dispersion corresponding to each of the regional coordinate distribution subnetworks to output the set dispersion dispersion corresponding to the regional coordinate distribution subnetwork set;

[0069] For each of the regional coordinate distribution sub-network sets, a fusion calculation is performed on the coordinate quantity discreteness corresponding to the regional coordinate distribution sub-network set and the set dispersion discreteness corresponding to the regional coordinate distribution sub-network set (for example, a weighted mean calculation can be performed) to output the fused discreteness corresponding to the regional coordinate distribution sub-network set, and then a regional coordinate distribution sub-network set having the minimum corresponding fused discreteness is screened out from the multiple regional coordinate distribution sub-network sets to be marked as the target regional coordinate distribution sub-network set;

[0070] For each regional coordinate distribution sub-network included in the target regional coordinate distribution sub-network set, determining whether the original path regions corresponding to each two regional four-dimensional coordinates in the regional coordinate distribution sub-network have a target region attribute, and if the target region attribute is present, filtering out one of the two corresponding original path regions (for example, the one with a smaller region area or path length may be filtered out), the target region attribute being used to reflect that the two corresponding original path regions have an adjacent relationship on the target navigation path;

[0071] Each original path region that is not filtered out is marked as a path region.

[0072] For example, in some more detailed implementations, in order to implement step S120, the following achievable sub-steps may be further performed:

[0073] The target vehicle air-conditioning environment data is loaded into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, so as to perform energy consumption prediction processing on the target vehicle air-conditioning environment data using the target air-conditioning energy consumption prediction neural network, outputting first matching probabilities corresponding to the target vehicle air-conditioning environment data in a plurality of initial energy consumption label information, and outputting updated second matching probabilities corresponding to the target vehicle air-conditioning environment data in at least one extended energy consumption label information (the target vehicle air-conditioning environment data has a first matching probability corresponding to each initial energy consumption label information, the first matching probability reflecting the predicted output of the target vehicle air-conditioning environment data in the initial energy consumption label information. The target vehicle air-conditioning environment data has a second matching probability corresponding to each extended energy consumption label information, the second matching probability reflecting the updated predicted output of the target vehicle air-conditioning environment data in the extended energy consumption label information). The plurality of initial energy consumption label information and the at least one extended energy consumption label information are respectively configured based on exemplary vehicle air-conditioning environment data of different time periods, and the exemplary vehicle air-conditioning environment data corresponding to the initial energy consumption label information is generated earlier than the exemplary vehicle air-conditioning environment data corresponding to the extended energy consumption label information, and each of the initial energy consumption label information and each of the initial energy consumption label information are different.

[0074] Based on the first matching probabilities corresponding to the multiple initial energy consumption label information and the second matching probabilities corresponding to the at least one extended energy consumption label information, the matching energy consumption label information corresponding to the target automobile air-conditioning environment data is matched from the multiple initial energy consumption label information and the at least one extended energy consumption label information, and then the energy consumption value reflected by the matching energy consumption label information is marked as the corresponding target energy consumption prediction result (for example, based on the first matching probabilities corresponding to the multiple initial energy consumption label information and the second matching probabilities corresponding to the at least one extended energy consumption label information, the target matching probabilities of the target automobile air-conditioning environment data corresponding to the at least one extended energy consumption label information and the multiple initial energy consumption label information can be calculated. Then, the energy consumption label information with the largest matching probability, that is, the extended energy consumption label information or the initial energy consumption label information, is used as the matching energy consumption label information corresponding to the target automobile air-conditioning environment data).

[0075] For example, in some more detailed implementations, in order to implement the steps of loading the target automobile air-conditioning environment data into the target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, performing energy consumption prediction processing on the target automobile air-conditioning environment data using the target air-conditioning energy consumption prediction neural network, outputting first matching probabilities corresponding to the target automobile air-conditioning environment data in multiple initial energy consumption label information, and outputting updated second matching probabilities corresponding to the target automobile air-conditioning environment data in at least one extended energy consumption label information, the following achievable sub-steps may be further performed:

[0076] The target automobile air-conditioning environment data is loaded into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, and a data mining operation is performed on the target automobile air-conditioning environment data using the target air-conditioning energy consumption prediction neural network (feature mining can be performed using a convolutional network included in the target air-conditioning energy consumption prediction neural network), so as to output a corresponding first environment data mining result; and based on the first environment data mining result, an analysis is output (probabilistic analysis can be performed using a softmax function) of first matching probabilities corresponding to the target automobile air-conditioning environment data in the at least one extended energy consumption label information and the plurality of initial energy consumption label information;

[0077] The target air-conditioning energy consumption prediction neural network is utilized, and the first matching probability corresponding to the at least one extended energy consumption label information is updated based on the reference probability update parameter to output the second matching probability corresponding to the at least one extended energy consumption label information (for example, the sum of the first matching probability and the reference probability update parameter can be used as the second matching probability; in this way, after obtaining the first matching probability corresponding to the at least one extended energy consumption label information of the target automobile air-conditioning environment data, the target air-conditioning energy consumption prediction neural network updates the first matching probability corresponding to the at least one extended energy consumption label information by using the reference probability update parameter, and a second matching probability with higher accuracy can be obtained, thereby improving the accuracy of the prediction when the second matching probability is used to determine the energy consumption corresponding to the target automobile air-conditioning environment data).

[0078] For example, in some more detailed embodiments, the automobile air conditioning energy consumption prediction method may further include a step of optimizing the target air conditioning energy consumption prediction neural network. To implement this step, the following achievable sub-steps may be further performed:

[0079] Using the extracted first and second example environment data sets, a network optimization operation is performed on the initial air conditioning energy consumption prediction neural network, so as to output an intermediate air conditioning energy consumption prediction neural network when a learning cost value formed during the network optimization process is less than a pre-configured learning cost value threshold. The first example environment data set corresponds to at least one extended energy consumption label information (exemplarily, the extended energy consumption label information is used to reflect the actual unit energy consumption value corresponding to the exemplary extended automobile air conditioning environment data included in the first example environment data set), and the second example environment data set corresponds to multiple initial energy consumption label information (exemplarily, the initial energy consumption label information is used to reflect the actual unit energy consumption value corresponding to the exemplary initial automobile air conditioning environment data included in the second example environment data set; and exemplarily, the total number of updated exemplary automobile air conditioning environment data corresponding to the at least one extended energy consumption label information is the same as the total number of updated exemplary automobile air conditioning environment data corresponding to the multiple initial energy consumption label information).

[0080] Extracting multiple updated exemplary automobile air-conditioning environment data, then loading each of the updated exemplary automobile air-conditioning environment data into the intermediate air-conditioning energy consumption prediction neural network to perform energy consumption prediction processing using the intermediate air-conditioning energy consumption prediction neural network, outputting a matching probability of the first extended energy consumption label information corresponding to the corresponding updated exemplary automobile air-conditioning environment data in the at least one extended energy consumption label information, and outputting a matching probability of the first initial energy consumption label information corresponding to the corresponding updated exemplary automobile air-conditioning environment data in the multiple initial energy consumption label information, each of the updated exemplary automobile air-conditioning environment data corresponding to one of the extended energy consumption label information or one of the initial energy consumption label information (that is, the actual unit energy consumption value corresponding to the updated exemplary automobile air-conditioning environment data is either the same as the actual unit energy consumption value corresponding to one of the extended energy consumption label information or the same as the actual unit energy consumption value corresponding to one of the initial energy consumption label information, and cannot be a unit energy consumption value other than the actual unit energy consumption value corresponding to the extended energy consumption label information and the actual unit energy consumption value corresponding to the initial energy consumption label information);

[0081] Based on the predicted matching probability of the first extended energy consumption label information and the first initial energy consumption label information, the corresponding reference probability update parameter is analyzed and output, and then the reference probability update parameter is used to update the output of the intermediate air-conditioning energy consumption prediction neural network to form the corresponding target air-conditioning energy consumption prediction neural network (exemplarily, updating the output of the intermediate air-conditioning energy consumption prediction neural network based on the reference probability update parameter may refer to updating the extended energy consumption label information matching probability corresponding to at least one extended energy consumption label information predicted and output by the intermediate air-conditioning energy consumption prediction neural network based on the same reference probability update parameter to obtain the updated extended energy consumption label information matching probability; it may also refer to updating the initial energy consumption label information matching probability corresponding to multiple initial energy consumption label information predicted and output by the intermediate air-conditioning energy consumption prediction neural network based on the same reference probability update parameter to obtain the updated initial energy consumption label information matching probability).

[0082] Because the multiple initial energy consumption tag information and the at least one extended energy consumption tag information are configured based on exemplary vehicle air conditioning environment data for different time periods, the initial air conditioning energy consumption prediction neural network is updated and optimized based on the exemplary vehicle air conditioning environment data for different time periods to form an intermediate air conditioning energy consumption prediction neural network. This intermediate air conditioning energy consumption prediction neural network learns the prediction and matching capabilities corresponding to the extended energy consumption tag information while also not possessing the prediction and matching capabilities corresponding to the initial energy consumption tag information. Furthermore, the intermediate air conditioning energy consumption prediction neural network performs energy consumption prediction processing on the multiple updated exemplary vehicle air conditioning environment data. Based on the predicted matching probabilities of the multiple updated exemplary vehicle air conditioning environment data for the energy consumption tag information for the two time periods, corresponding reference probability update parameters are analyzed. The reference probability update parameters are then used to update the output of the intermediate air conditioning energy consumption prediction neural network to form a target air conditioning energy consumption prediction neural network. This ensures that the matching probabilities output by the target air conditioning energy consumption prediction neural network for the energy consumption tag information for the two time periods are balanced, thereby improving the accuracy of the neural network prediction. Furthermore, directly using the reference probability update parameters to update the output of the intermediate air conditioning energy consumption prediction neural network eliminates the need for repeated training, thereby improving efficiency and reducing resource consumption.

[0083] For example, in some more detailed embodiments, in order to implement the step of performing a network optimization operation on the initial air conditioning energy consumption prediction neural network using the extracted first example environment data set and the second example environment data set, and outputting an intermediate air conditioning energy consumption prediction neural network when a learning cost value formed during the network optimization process is less than a pre-configured learning cost value threshold, the following achievable sub-steps may be further performed:

[0084] Each exemplary extended automobile air-conditioning environment data in the first exemplary environment data set is loaded into the initial air-conditioning energy consumption prediction neural network, so as to use the initial air-conditioning energy consumption prediction neural network for analysis and prediction processing, and output the matching probability of the corresponding exemplary extended automobile air-conditioning environment data in the at least one extended energy consumption label information corresponding to the second extended energy consumption label information (one exemplary extended automobile air-conditioning environment data corresponds to one second extended energy consumption label information matching probability), and output the matching probability of the corresponding exemplary extended automobile air-conditioning environment data in the multiple initial energy consumption label information corresponding to the second initial energy consumption label information (one exemplary extended automobile air-conditioning environment data corresponds to one second initial energy consumption label information). Initial energy consumption label information matching probability; in addition, if the at least one extended energy consumption label information is respectively extended energy consumption label information 1 and extended energy consumption label information 2, the multiple initial energy consumption label information are respectively initial energy consumption label information 1, initial energy consumption label information 2, and initial energy consumption label information 3, and the actual energy consumption value corresponding to the exemplary extended automobile air-conditioning environment data A is extended energy consumption label information 1, then the actual matching probability corresponding to the exemplary extended automobile air-conditioning environment data A on extended energy consumption label information 1 is 1, the actual matching probability corresponding to extended energy consumption label information 2 is 0, the actual matching probability on initial energy consumption label information 1 is 0, the actual matching probability on initial energy consumption label information 2 is 0, and the actual matching probability on initial energy consumption label information 3 is 0);

[0085] Loading each exemplary initial automobile air-conditioning environment data in the second exemplary environment data set into the initial air-conditioning energy consumption prediction neural network, performing analysis and prediction processing using the initial air-conditioning energy consumption prediction neural network, outputting a matching probability of the corresponding exemplary initial automobile air-conditioning environment data with the third extended energy consumption label information corresponding to the at least one extended energy consumption label information (exemplarily, one exemplary initial automobile air-conditioning environment data corresponds to one third extended energy consumption label information matching probability), and outputting a matching probability of the corresponding exemplary initial automobile air-conditioning environment data with the third initial energy consumption label information corresponding to the multiple initial energy consumption label information (exemplarily, one exemplary initial automobile air-conditioning environment data corresponds to one third initial energy consumption label information matching probability);

[0086] Based on the analysis output of the second extended energy consumption label information matching probability, the second initial energy consumption label information matching probability, the third extended energy consumption label information matching probability and the third initial energy consumption label information matching probability, the corresponding network optimization learning cost value is calculated and output, and then the network optimization operation is performed on the initial air-conditioning energy consumption prediction neural network based on the network optimization learning cost value. When the learning cost value formed during the network optimization process is less than the pre-configured learning cost value threshold (that is, the convergence condition is met), the intermediate air-conditioning energy consumption prediction neural network corresponding to the initial air-conditioning energy consumption prediction neural network is output.

[0087] For example, in some more detailed implementations, in order to implement the step of calculating and outputting corresponding network optimization learning costs based on the second extended energy consumption label information matching probability, the second initial energy consumption label information matching probability, the third extended energy consumption label information matching probability, and the third initial energy consumption label information matching probability outputted from the analysis, the following achievable sub-steps may be further performed:

[0088] Calculate and output a first-category learning cost based on the analyzed matching probability of the second extended energy label information, the matching probability of the second initial energy label information, the matching probability of the third extended energy label information, and the matching probability of the third initial energy label information (the degree of difference between the matching probability and the actual matching probability can be used as the first-category learning cost);

[0089] Based on the second initial energy consumption label information matching probability and the third initial energy consumption label information matching probability outputted from the analysis, a second type of learning cost value is calculated and outputted (exemplarily, the initial air-conditioning energy consumption prediction neural network can be constructed and formed based on the relevant air-conditioning energy consumption prediction neural network and the corresponding relevant reference probability update parameters. The exemplary automobile air-conditioning environment data corresponding to the relevant air-conditioning energy consumption prediction neural network can be formed earlier. Based on this, energy consumption prediction processing can be performed on the exemplary extended automobile air-conditioning environment data and the exemplary initial automobile air-conditioning environment data respectively based on the relevant air-conditioning energy consumption prediction neural network to output the corresponding first relevant initial energy consumption label information matching probability and the second relevant initial energy consumption label information matching probability. The second type of learning cost value is then determined based on the degree of difference between the second initial energy consumption label information matching probability and the third initial energy consumption label information matching probability and the first relevant initial energy consumption label information matching probability and the second relevant initial energy consumption label information matching probability);

[0090] The first type of learning cost value and the second type of learning cost value are integrated (such as weighted sum calculation, etc.), and the corresponding network optimization learning cost value is output.

[0091] For example, in some more detailed implementations, in order to implement the step of extracting a plurality of updated exemplary automobile air-conditioning environment data, the following achievable sub-steps may be performed:

[0092] Multiple updated exemplary automobile air-conditioning environment data are extracted from the first example environment data set and the second example environment data set, respectively; and / or multiple updated exemplary automobile air-conditioning environment data are extracted from a third example environment data set other than the first example environment data set and the second example environment data set.

[0093] For example, in some more detailed implementations, in order to achieve the step of analyzing and outputting the corresponding reference probability update parameters based on the predicted output of the first extended energy consumption label information and the first initial energy consumption label information, and then using the reference probability update parameters to update the output of the intermediate air conditioning energy consumption prediction neural network to form the corresponding target air conditioning energy consumption prediction neural network, the following achievable sub-steps may be further performed:

[0094] Calculating and outputting an extended energy label representative value corresponding to the at least one extended energy label information based on the predicted matching probability of the first extended energy label information (for example, an average value of the matching probabilities of the first extended energy label information may be first calculated, and then the average value may be used as the extended energy label representative value);

[0095] Based on the predicted matching probabilities of the first initial energy consumption label information, calculate and output initial energy consumption label representative values ​​corresponding to the multiple initial energy consumption label information (for example, an average value of the matching probabilities of the first initial energy consumption label information may be first calculated, and then the average value may be used as the initial energy consumption label representative value);

[0096] Based on the extended energy consumption label representative value and the initial energy consumption label representative value, calculating and outputting a corresponding reference probability update parameter (exemplarily, the reference probability update parameter may be obtained by performing a difference calculation on the extended energy consumption label representative value and the initial energy consumption label representative value);

[0097] The reference probability update parameter is configured in the output subnetwork of the intermediate air-conditioning energy consumption prediction neural network to form a corresponding target air-conditioning energy consumption prediction neural network, and the reference probability update parameter is used to update the extended energy consumption label information matching probability corresponding to the at least one extended energy consumption label information (exemplarily, the update may refer to summing the extended energy consumption label information matching probability and the reference probability update parameter).

[0098] For example, in some more detailed implementations, in order to implement step S130, the following achievable sub-steps may be further performed:

[0099] Extracting the actual air conditioning energy consumption value of the target vehicle air conditioner within a recent valid time period based on a pre-configured valid time period (e.g., one week or one month), where the actual air conditioning energy consumption value reflects the unit energy consumption of the target vehicle air conditioner within the recent time period (for example, the unit energy consumption may refer to the average hourly energy consumption);

[0100] The actual air-conditioning energy consumption value and the unit predicted energy consumption reflected by the target energy consumption prediction result (the unit predicted energy consumption can also refer to the average of the predicted hourly energy consumption) are weightedly summed up (the specific weight of the weighted summation can be configured according to actual application requirements) to output the target energy consumption fusion result corresponding to the target automobile air-conditioning.

[0101] Combine Figure 3 The embodiment of the present invention further provides an automobile air conditioning energy consumption prediction device, which can be applied to the above automobile air conditioning energy consumption prediction system. The automobile air conditioning energy consumption prediction device may include:

[0102] An air-conditioning environment data acquisition module is used to collect target automobile air-conditioning environment data corresponding to a target automobile air-conditioning, wherein the target automobile air-conditioning environment data is used to reflect the environment of the space where the target automobile air-conditioning is located;

[0103] an energy consumption prediction processing module, configured to load the target automobile air-conditioning environment data into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, perform energy consumption prediction processing on the target automobile air-conditioning environment data using the target air-conditioning energy consumption prediction neural network, and output a corresponding target energy consumption prediction result;

[0104] The energy consumption prediction fusion module is used to extract the actual air conditioning energy consumption value of the target automobile air conditioner in the most recent time period, and fuse the actual air conditioning energy consumption value with the target energy consumption prediction result to output the target energy consumption fusion result corresponding to the target automobile air conditioner.

[0105] In summary, the present invention provides a method and system for predicting the energy consumption of an automobile air conditioner. The method collects target automobile air conditioner environmental data corresponding to a target automobile air conditioner; loads the target automobile air conditioner environmental data into a target air conditioner energy consumption prediction neural network that has undergone an optimized operation, and uses the target air conditioner energy consumption prediction neural network to perform energy consumption prediction processing on the target automobile air conditioner environmental data, and outputs a corresponding target energy consumption prediction result; extracts the actual air conditioning energy consumption value of the target automobile air conditioner in a recent time period, and fuses the actual air conditioning energy consumption value with the target energy consumption prediction result to output a target energy consumption fusion result corresponding to the target automobile air conditioner. Based on this, when predicting and analyzing air conditioning energy consumption, not only the historical energy consumption of the automobile air conditioner itself will be considered, but also the results of predictive analysis of environmental data based on the neural network will be considered, realizing the fusion of multi-dimensional data, making the basis for predictive analysis sufficient, thereby improving the reliability of automobile air conditioner energy consumption prediction to a certain extent.

[0106] 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 are intended to be within the scope of protection of the present invention.

Claims

1. A method for predicting energy consumption of automobile air conditioner, characterized in that: include: collecting target automobile air-conditioning environment data corresponding to the target automobile air-conditioning, wherein the target automobile air-conditioning environment data is used to reflect the environment of the space where the target automobile air-conditioning is located; The target automobile air-conditioning environment data is loaded into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, so as to perform energy consumption prediction processing on the target automobile air-conditioning environment data using the target air-conditioning energy consumption prediction neural network and output a corresponding target energy consumption prediction result; The actual air conditioning energy consumption value of the target automobile air conditioner in a recent time period is extracted, and the actual air conditioning energy consumption value and the target energy consumption prediction result are fused to output a target energy consumption fusion result corresponding to the target automobile air conditioner.

2. The automobile air conditioning energy consumption prediction method according to claim 1, characterized in that: The step of collecting target automobile air-conditioning environment data corresponding to the target automobile air-conditioning comprises: Extracting and processing the navigation data of the target vehicle corresponding to the target vehicle air conditioner to output target navigation data corresponding to the target vehicle, and then performing data analysis and processing on the target navigation data to determine a target navigation path corresponding to the target navigation data; Environmental data collection and processing are performed on each path area on the target navigation path to form target automobile air-conditioning environmental data corresponding to the target automobile air-conditioning.

3. The automobile air conditioning energy consumption prediction method according to claim 1, characterized in that: The step of loading the target automobile air-conditioning environment data into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, performing energy consumption prediction processing on the target automobile air-conditioning environment data using the target air-conditioning energy consumption prediction neural network, and outputting a corresponding target energy consumption prediction result includes: The target automobile air-conditioning environment data is loaded into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, so as to use the target air-conditioning energy consumption prediction neural network to perform energy consumption prediction processing on the target automobile air-conditioning environment data, output first matching probabilities corresponding to the target automobile air-conditioning environment data in a plurality of initial energy consumption label information, and output updated second matching probabilities corresponding to the target automobile air-conditioning environment data in at least one extended energy consumption label information, wherein the plurality of initial energy consumption label information and the at least one extended energy consumption label information are respectively configured based on exemplary automobile air-conditioning environment data of different time periods, and the data formation time of the exemplary automobile air-conditioning environment data corresponding to the initial energy consumption label information is earlier than the data formation time of the exemplary automobile air-conditioning environment data corresponding to the extended energy consumption label information, and each of the initial energy consumption label information and each of the initial energy consumption label information are different; Based on the first matching probabilities corresponding to the multiple initial energy consumption label information and the second matching probabilities corresponding to the at least one extended energy consumption label information, the matching energy consumption label information corresponding to the target automobile air-conditioning environment data is matched from the multiple initial energy consumption label information and the at least one extended energy consumption label information, and then the energy consumption value reflected by the matching energy consumption label information is marked as the corresponding target energy consumption prediction result.

4. The automobile air conditioning energy consumption prediction method according to claim 3, characterized in that: The step of loading the target automobile air-conditioning environment data into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, performing energy consumption prediction processing on the target automobile air-conditioning environment data using the target air-conditioning energy consumption prediction neural network, outputting first matching probabilities corresponding to the target automobile air-conditioning environment data in a plurality of initial energy consumption label information, and outputting updated second matching probabilities corresponding to the target automobile air-conditioning environment data in at least one extended energy consumption label information, includes: The target vehicle air-conditioning environment data is loaded into a target air-conditioning energy consumption prediction neural network that has undergone an optimized operation, and a data mining operation is performed on the target vehicle air-conditioning environment data using the target air-conditioning energy consumption prediction neural network to output a corresponding first environment data mining result, and then, based on the first environment data mining result, a first matching probability corresponding to the target vehicle air-conditioning environment data in the at least one extended energy consumption label information and the plurality of initial energy consumption label information is analyzed and output; The target air conditioner energy consumption prediction neural network is used, and based on the reference probability update parameter, the first matching probability corresponding to the at least one extended energy consumption label information is updated to output the second matching probability corresponding to the at least one extended energy consumption label information.

5. The automobile air conditioning energy consumption prediction method according to claim 3, characterized in that: The automobile air-conditioning energy consumption prediction method further includes the step of optimizing the target air-conditioning energy consumption prediction neural network, which includes: Using the extracted first example environment data set and the second example environment data set, a network optimization operation is performed on the initial air conditioning energy consumption prediction neural network, so as to output an intermediate air conditioning energy consumption prediction neural network when a learning cost value formed during the network optimization process is less than a pre-configured learning cost value threshold, wherein the first example environment data set corresponds to at least one extended energy consumption label information, and the second example environment data set corresponds to multiple initial energy consumption label information; Extracting multiple updated exemplary automobile air-conditioning environment data, then loading each of the updated exemplary automobile air-conditioning environment data into the intermediate air-conditioning energy consumption prediction neural network, using the intermediate air-conditioning energy consumption prediction neural network to perform energy consumption prediction processing, outputting a matching probability of the first extended energy consumption label information corresponding to the corresponding updated exemplary automobile air-conditioning environment data in the at least one extended energy consumption label information, and outputting a matching probability of the first initial energy consumption label information corresponding to the corresponding updated exemplary automobile air-conditioning environment data in the multiple initial energy consumption label information, each of the updated exemplary automobile air-conditioning environment data corresponding to one of the extended energy consumption label information or one of the initial energy consumption label information; Based on the predicted matching probability of the first extended energy consumption label information and the first initial energy consumption label information, the corresponding reference probability update parameters are analyzed and output, and then the reference probability update parameters are used to update the output of the intermediate air-conditioning energy consumption prediction neural network to form the corresponding target air-conditioning energy consumption prediction neural network.

6. The automobile air conditioning energy consumption prediction method according to claim 5, characterized in that: The step of analyzing and outputting corresponding reference probability update parameters based on the predicted matching probability of the first extended energy consumption label information and the first initial energy consumption label information, and then using the reference probability update parameters to update the output of the intermediate air conditioning energy consumption prediction neural network to form a corresponding target air conditioning energy consumption prediction neural network includes: Calculate and output an extended energy consumption label representative value corresponding to the at least one extended energy consumption label information based on the predicted matching probability of the first extended energy consumption label information; Calculate and output initial energy consumption label representative values ​​corresponding to the plurality of initial energy consumption label information based on the predicted matching probability of the first initial energy consumption label information; Based on the extended energy consumption label representative value and the initial energy consumption label representative value, calculating and outputting a corresponding reference probability update parameter; The reference probability update parameter is configured in the output subnetwork of the intermediate air conditioning energy consumption prediction neural network to form a corresponding target air conditioning energy consumption prediction neural network, and the reference probability update parameter is used to update the matching probability of the extended energy consumption label information corresponding to the at least one extended energy consumption label information.

7. The automobile air conditioning energy consumption prediction method according to claim 5, characterized in that: The step of performing a network optimization operation on the initial air conditioning energy consumption prediction neural network using the extracted first example environment data set and the second example environment data set, so as to output an intermediate air conditioning energy consumption prediction neural network when a learning cost value formed during the network optimization process is less than a pre-configured learning cost value threshold, includes: Loading each exemplary extended automobile air-conditioning environment data in the first exemplary environment data set into an initial air-conditioning energy consumption prediction neural network, performing analysis and prediction processing using the initial air-conditioning energy consumption prediction neural network, outputting a matching probability of the corresponding exemplary extended automobile air-conditioning environment data with the second extended energy consumption label information corresponding to the at least one extended energy consumption label information, and outputting a matching probability of the corresponding exemplary extended automobile air-conditioning environment data with the second initial energy consumption label information corresponding to the multiple initial energy consumption label information; Loading each exemplary initial automobile air-conditioning environment data in the second exemplary environment data set into the initial air-conditioning energy consumption prediction neural network, performing analysis and prediction processing using the initial air-conditioning energy consumption prediction neural network, outputting a matching probability of the corresponding exemplary initial automobile air-conditioning environment data with the third extended energy consumption label information corresponding to the at least one extended energy consumption label information, and outputting a matching probability of the corresponding exemplary initial automobile air-conditioning environment data with the third initial energy consumption label information corresponding to the multiple initial energy consumption label information; Based on the analysis output of the second extended energy consumption label information matching probability, the second initial energy consumption label information matching probability, the third extended energy consumption label information matching probability and the third initial energy consumption label information matching probability, the corresponding network optimization learning cost value is calculated and output, and then the network optimization operation is performed on the initial air-conditioning energy consumption prediction neural network based on the network optimization learning cost value, and when the learning cost value formed during the network optimization process is less than the pre-configured learning cost value threshold, the intermediate air-conditioning energy consumption prediction neural network corresponding to the initial air-conditioning energy consumption prediction neural network is output.

8. The automobile air-conditioning energy consumption prediction method according to claim 7, characterized in that: The step of calculating and outputting a corresponding network optimization learning cost value based on the analyzed second extended energy consumption label information matching probability, the second initial energy consumption label information matching probability, the third extended energy consumption label information matching probability, and the third initial energy consumption label information matching probability includes: Calculate and output a first-category learning cost based on the second extended energy consumption label information matching probability, the second initial energy consumption label information matching probability, the third extended energy consumption label information matching probability, and the third initial energy consumption label information matching probability outputted from the analysis; Calculate and output a second type of learning cost value based on the second initial energy consumption label information matching probability and the third initial energy consumption label information matching probability outputted from the analysis; The first type of learning cost value and the second type of learning cost value are integrated to output a corresponding network optimization learning cost value.

9. The automobile air-conditioning energy consumption prediction method according to any one of claims 1 to 8, characterized in that: The step of extracting the actual air conditioning energy consumption value of the target automobile air conditioner in a recent time period, and fusing the actual air conditioning energy consumption value with the target energy consumption prediction result to output a target energy consumption fusion result corresponding to the target automobile air conditioner includes: Extracting, based on a pre-configured effective time length, an actual air conditioning energy consumption value of the target automobile air conditioner in a most recent effective time period, the actual air conditioning energy consumption value being used to reflect the unit energy consumption of the target automobile air conditioner in the most recent time period; The actual air-conditioning energy consumption value and the unit predicted energy consumption reflected by the target energy consumption prediction result are weighted and calculated to output a target energy consumption fusion result corresponding to the target automobile air-conditioning.

10. An automobile air conditioning energy consumption prediction system, characterized in that: The system comprises 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 automobile air conditioning energy consumption prediction method according to any one of claims 1 to 9.

Images