A method and system for implementing customer service of a heating system based on complex demand analysis

Abstract

The present invention relates to a method and system for implementing customer service of a heating system based on complex demand analysis, and relates to the technical field of customer service management, including: based on a heating customer service platform, collecting object feature information of multiple heat-using objects in a heating system, performing heat priority analysis, and obtaining multiple heat priorities; screening sample data according to multiple heat priorities to obtain a heating demand analysis data set; using a heating demand analysis data set according to multiple heat priorities to train a heating demand analyzer; collecting current heat demand information of multiple heat-using objects, integrating and analyzing, and obtaining multiple heating parameters; setting multiple control orders, performing impact analysis and correction on multiple heating parameters, obtaining multiple corrected heating parameters, and performing customer feedback and heating control. The present invention achieves the technical effect of improving the efficiency and accuracy of complex heating demand analysis.

Description

Technical Field

[0001] The present invention relates to the field of customer service management, and in particular to a method and system for implementing customer service of a heating system based on complex demand analysis. Background Art

[0002] In modern urban heating systems, the effectiveness of heating services and user satisfaction largely depend on the accurate analysis and response of heating demand. The heating system needs to deal with multiple heat users, such as multiple households, and different heat users have complex and diverse heating demands. This diversity places higher demands on the heating system, requiring the customer service of the heating platform to not only respond to heating demands in real time, but also accurately analyze and dispatch different heating demands.

[0003] The current analysis of heating demand mainly relies on historical data and empirical rules. Although these methods can meet the needs to a certain extent, due to the dynamic changes in heating demand and the complex influence of the external environment, after collecting heating demand through the heating platform customer service, the analysis accuracy and efficiency of complex heating demand of multiple heat users are low, affecting the quality of heating service under complex demands of multiple heat users. Summary of the invention

[0004] The present invention aims to solve the technical problem that the heating customer service system in the prior art has low accuracy and efficiency in analyzing the complex heating demands of multiple heat-using objects, which affects the quality of heating services under the complex demands of multiple heat-using objects.

[0005] In a first aspect, the present invention provides a method for implementing customer service of a heating system based on complex demand analysis, including: based on a heating customer service platform, collecting object feature information of multiple heat-using objects in the heating system, performing heat-using priority analysis, and obtaining multiple heat-using priorities;

[0006] Collecting historical heat consumption data of the plurality of heat-consuming objects, and screening sample data according to the plurality of heat consumption priorities to obtain a heat demand analysis data set, wherein the heat demand analysis data set includes a plurality of object demand analysis data sets;

[0007] According to the multiple heat usage priorities, the heat demand parsing data set is used to train a heat demand parser, wherein initial data weights are set according to the multiple heat usage priorities, and the data weights are adjusted during training according to the training results;

[0008] Based on the heating customer service platform, current heating demand information of the plurality of heating users is collected, and integrated analysis is performed based on the heating demand analyzer to obtain a plurality of heating parameters;

[0009] Based on the multiple heat usage priorities, multiple control orders are set, and the multiple heating parameters are impact analyzed and corrected to obtain multiple corrected heating parameters as the heating demand analysis result, and customer feedback and heating control are performed.

[0010] In a second aspect, the present invention provides a customer service implementation system for a heating system based on complex demand analysis, the system comprising: a priority analysis module for collecting object feature information of multiple heat-using objects in a heating system based on a heating customer service platform, performing heat-using priority analysis, and obtaining multiple heat-using priorities;

[0011] A data collection and screening module, used to collect historical heat consumption data of the plurality of heat consumption objects, and screen sample data according to the plurality of heat consumption priorities to obtain a heat demand analysis data set, wherein the heat demand analysis data set includes a plurality of object demand analysis data sets;

[0012] A parsing training module, configured to train a heat demand parser using the heat demand parsing data set according to the plurality of heat use priorities, wherein initial data weights are set according to the plurality of heat use priorities, and the data weights are adjusted during training according to training results;

[0013] A demand analysis module, for collecting the current heat demand information of the plurality of heat users based on the heat supply customer service platform, and performing integrated analysis based on the heat supply demand analyzer to obtain a plurality of heat supply parameters;

[0014] The heating feedback module is used to set multiple control orders based on the multiple heat usage priorities, perform impact analysis and correction on the multiple heating parameters, obtain multiple corrected heating parameters as the heating demand analysis results, and perform customer feedback and heating control.

[0015] In a third aspect, the present invention provides an electronic device comprising: a memory for storing a computer software program; and a processor for reading and executing the computer software program, thereby realizing the customer service implementation method of a heating system based on complex demand analysis of the first aspect.

[0016] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer software program stored therein, and when the computer software program is executed by a processor, the method for implementing customer service of a heating system based on complex demand analysis according to the first aspect is implemented.

[0017] The present invention collects object feature information of multiple heat-using objects in a heating system based on a heating customer service platform, performs heat priority analysis, obtains multiple heat priorities, and then collects historical heat usage data of multiple heat-using objects. Sample data is screened according to the multiple heat priorities to obtain a heating demand analysis data set, which can screen more heating demand analysis data of heat-using objects with larger priorities, thereby improving the accuracy of subsequent training of a heating demand analyzer for analyzing the heat demand of heat-using objects with larger priorities. Then, according to the multiple heat priorities, the heating demand analysis data set is used to train the heating demand analyzer, wherein initial data weights are set according to the multiple heat priorities, and the initial data weights are used during training. In the method, the data weight is adjusted according to the training results, the heating demand analysis data is randomly selected according to the weight, and the heating demand analyzer is trained by integrated learning, which can improve the overall accuracy of the heating demand analyzer. Based on the heating customer service platform, the current heating demand information of multiple heating objects is collected, and integrated analysis is performed based on the heating demand analyzer to obtain multiple heating parameters. Based on multiple heating priorities, multiple control orders are set, and the impact analysis and correction of multiple heating parameters are performed to obtain multiple corrected heating parameters as the heating demand analysis results, and customer feedback and heating control are performed to reduce the heating impact of neighboring heating objects on heating objects with higher priorities, thereby improving the quality of heating services. Through the above scheme, the present application considers the priorities of multiple heating objects, performs data collection and screening, and performs data weight setting and integrated training of the heating demand analyzer, which can improve the analysis accuracy and efficiency of the training heating demand analyzer, and the analysis accuracy of the demand with higher priority is greater, achieving the technical effect of improving the analysis accuracy and response efficiency of complex heating demands. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flow chart of a method for implementing customer service of a heating system based on complex demand analysis provided by an embodiment of the present invention;

[0019] Figure 2 A schematic diagram of the structure of a customer service implementation system for a heating system based on complex demand analysis provided by an embodiment of the present invention;

[0020] Figure 3 A functional block diagram of an electronic device provided by an embodiment of the present invention;

[0021] Figure 4 A principle block diagram of the computer-readable storage medium provided by the present invention.

[0022] In the accompanying drawings, the components represented by the reference numerals are listed as follows:

[0023] Priority analysis module 11, data collection and screening module 12, analysis and training module 13, demand analysis module 14, heating feedback module 15, electronic device 100, memory 110, processor 120, computer program 111, computer readable storage medium 200, computer program 211. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0025] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, "plurality" means two or more, unless otherwise clearly and specifically defined.

[0026] In the description of the present application, the term "for example" is used to mean "used as an example, illustration or description". Any embodiment described as "for example" in the present application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid unnecessary details to obscure the description of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present application.

[0027] Example 1

[0028] like Figure 1 As shown, the embodiment of the present invention provides a method for implementing customer service of a heating system based on complex demand analysis, including:

[0029] S10: Based on the heating customer service platform, object feature information of multiple heat-using objects in the heating system is collected, and heat-using priority analysis is performed to obtain multiple heat-using priorities;

[0030] In the embodiment of the present application, in the heating system, based on the heating customer service platform, the object feature information of multiple heat users is first comprehensively collected. The heating customer service platform can be a customer service platform built for the heating system, which is used to collect the heating needs of the heat users, i.e., users, for heating, and solve other heating problems.

[0031] The object feature information is key data related to heat demand, including but not limited to building area, historical heat consumption, historical heat consumption time, etc. These feature information provides necessary data support for subsequent heat priority analysis.

[0032] Analyze the collected object feature information and evaluate the priority of the heat demand of each heat-using object. Heat priority refers to the priority level of heat supply service for heat-using objects. The evaluation basis of this priority includes factors such as the building area, historical heat consumption, and historical heat consumption time of the object. For example, users with larger building areas, larger historical heat consumption, and longer historical heat consumption time need higher maintenance priority and have higher priority.

[0033] Through the above steps, the heat priority of each heat user can be obtained, which lays the foundation for the subsequent heat demand analysis and heat control. Objects with higher priority will receive more attention in the heat demand analysis, ensuring the reliability and accuracy of the heat service.

[0034] Step S10 in the method provided in the embodiment of the present application includes:

[0035] Based on the heating customer service platform, multiple object feature information of multiple heat-using objects in the heating system is collected;

[0036] According to the plurality of object feature information, a priority classification table is input to perform classification to obtain a plurality of heat usage priorities, wherein the priority classification table includes an index relationship between a plurality of sample object feature information and a plurality of sample priorities.

[0037] In the embodiment of the present application, in order to accurately meet the diverse needs of multiple heat users, firstly, multiple object feature information of these heat users is collected through the heat supply customer service platform. For example, one or more of the building area, historical heat consumption, historical heat consumption time, etc. of multiple heat users are collected as multiple object feature information.

[0038] After obtaining multiple object feature information, this information is input into the priority classification table for classification processing. The priority classification table is a pre-established database that stores a large number of object feature information of sample objects and corresponding heat usage priorities. The priority classification table here includes an index relationship between multiple sample object feature information and corresponding sample priorities. By matching and comparing the collected object feature information with the sample data in the classification table, each heat-using object can be assigned to an appropriate priority category according to the similarity principle, such as the corresponding sample priority indexed by the closest sample object feature information. Exemplarily, the priority category includes 10 priority levels from 1 to 10, and an index relationship between multiple sample historical heat usage and priority levels is constructed in the priority classification table, and the greater the sample historical heat usage, the greater the priority level of the index.

[0039] Optionally, the corresponding heat usage priority may be obtained based on the classification of object feature information based on other methods.

[0040] Through classification processing, the specific heating priority of each heat user can be obtained. This priority information provides an important reference for subsequent heating demand analysis, resource allocation and heating control, ensuring that high-priority objects receive more accurate heating demand analysis in heating services, thereby improving the quality and efficiency of overall heating services.

[0041] S20: collecting historical heat consumption data of the plurality of heat consumption objects, and screening sample data according to the plurality of heat consumption priorities to obtain a heat demand analysis data set, wherein the heat demand analysis data set includes a plurality of object demand analysis data sets;

[0042] In the embodiment of the present application, in the heating system, in order to realize the automatic analysis of heating demand, the historical heating data of multiple heating objects are first collected, and the historical heating data includes the heating demand information of multiple heating objects and the heating parameters that the heating objects consider qualified after the actual heating. The historical heating data of multiple heating objects are collected for subsequent training of a heating demand analyzer that automatically and accurately analyzes the user's heating demand.

[0043] The amount of historical heat usage data of multiple heat users is large. Direct training with such data has low convergence efficiency and slow training speed. Therefore, it is necessary to screen the historical heat usage data of multiple heat users. According to multiple heat usage priorities, sample data of the historical heat usage data of multiple heat users are screened to obtain a heat demand analysis data set, which includes multiple object demand analysis data sets of multiple heat users.

[0044] Among them, in the process of sample data screening, heat-using objects with higher priorities will be given greater weights. This means that in the screening process, more attention will be paid to the historical heat-using data of these objects to ensure that their heat demand can be parsed more accurately, to improve the parsing accuracy of users with high priority in complex multi-object heat demand parsing, and to improve the maintenance quality and heat service quality of users with higher priority. For example, in multiple object demand parsing data sets screened according to multiple heat priority levels, the greater the heat priority, the greater the amount of data in the object demand parsing data set for the application heat object.

[0045] Step S20 in the method provided in the embodiment of the present application includes:

[0046] According to the historical heat consumption data of the plurality of heat users stored in the heat supply customer service platform, a plurality of sample heat demand information sets are collected, and qualified heating parameters under different sample heat demand information are obtained to obtain a plurality of sample heating parameter sets;

[0047] Combining the plurality of sample heat demand information sets and the plurality of sample heating parameter sets to obtain a plurality of basic object demand analysis data sets;

[0048] According to the multiple heat usage priorities, a preset training data quantity N is allocated to obtain multiple object data quantities, the multiple basic object demand analysis data sets are randomly screened to obtain multiple object demand analysis data sets, and the heating demand analysis data set is integrated, where N is an integer greater than 1.

[0049] In the embodiment of the present application, the historical heat consumption data of multiple heat users stored in the heat supply customer service platform are first used. By analyzing these historical data, a set of sample heat demand information of multiple heat users can be collected. These sets include the actual heat demand of different heat users under various conditions, for example, the changes in heat demand under different temperatures, weather conditions, and the ages of people using heat at home, such as specific sample heat demand information such as "the temperature is a bit high", "the temperature is too high", "there are elderly people at home, the temperature is a bit low", etc.

[0050] Furthermore, qualified heating parameters that are considered qualified by the heat users after the heating adjustment is performed under these sample heat demand information are obtained. These heating parameters usually include heating temperature, heating pressure, heating time period, etc., which are heating parameters that ensure that the heat users achieve the expected results under different demands, such as heating parameters with a heating temperature of 20°C. By matching the sample heat demand information with its corresponding qualified heating parameters, multiple sample heat demand parameter sets can be formed. Multiple sample heat demand parameter sets provide basic data for subsequent demand analysis, ensuring that the optimal heating parameters under different heat demand information can be identified.

[0051] After obtaining multiple sample heat demand information sets and multiple sample heating parameter sets, they are combined to generate multiple basic object demand analysis data sets of the collection combination.

[0052] In order to improve the convergence efficiency of subsequent training of the heating demand parser, the preset number of training data N is set, that is, the total number of training data N for the heating demand parser to be trained, which is an integer greater than 1. It can be set according to the specific training requirements of the heating demand parser, for example, 10,000 sets of training data. Each set of training data includes sample heat demand information and sample heating parameters.

[0053] According to the multiple heat usage priorities obtained from the previous analysis, the preset number of training data N is allocated. By allocating the number of object data according to the heat usage priority, it is ensured that the data of high-priority heat-using objects occupy a larger proportion in the training set. This allocation strategy can improve the sensitivity and accuracy of the heating demand parser to the demand analysis of key heat-using objects. Exemplarily, the ratio of the heat usage priority of a certain heat-using object to the sum of multiple heat usage priorities is calculated, and the ratio is multiplied by the preset number of training data N as the number of object data allocated for the heat-using object, so as to obtain multiple numbers of object data.

[0054] According to the number of multiple object data, the multiple basic object demand analysis data sets are randomly screened to select demand analysis data to obtain multiple object demand analysis data sets. Random screening refers to randomly selecting sample data suitable for training under the premise of ensuring comprehensive data coverage. This method can avoid the problem of overfitting of data while ensuring the diversity and representativeness of training data. Finally, a heating demand analysis data set is generated by integrating multiple screened object demand analysis data sets. This data set will be used as training data for the heating demand parser to ensure that the heating demand parser can accurately and effectively respond to the heating demand of complex multiple heat-using objects in practical applications.

[0055] S30: training a heating demand parser using the heating demand parsing data set according to the plurality of heating priorities, wherein initial data weights are set according to the plurality of heating priorities, and the data weights are adjusted during training according to the training results;

[0056] In the embodiment of the present application, in order to improve the accuracy and efficiency of heat demand analysis, a heat demand analysis data set is used to train a heat demand analyzer based on multiple heat use priorities. Exemplarily, the heat demand analyzer is a model based on a machine learning algorithm, and its task is to accurately predict and analyze the specific heat demand of each heat user through analysis and learning of the input data, so as to obtain the corresponding heating parameters for heating.

[0057] Among them, according to the heat priority of each heat-using object, the initial data weight is set for the sample data in the heat demand analysis data set. Data weight refers to the degree of attention paid to a sample heat demand analysis data during the training process, such as the probability of being selected for training. For heat-using objects with higher priorities, their corresponding sample heat demand analysis data will be assigned a higher initial weight, which means that these data occupy a more important position in the training, enabling the trained heat demand parser to pay more attention to the needs of high-priority objects, improve the accuracy of heat demand analysis for high-priority heat-using objects, and ensure the quality of heat demand analysis for lower-priority heat-using objects.

[0058] This priority-based weight adjustment mechanism can effectively improve the learning ability of the heat demand parser for key data, making it better adapted to the complex demand analysis of multiple heat users. Through continuous weight adjustment, the heat demand parser can gradually optimize its performance, ensuring that the final generated model can accurately and efficiently analyze various complex heat demands, especially the heat demands of high-priority heat users, and improve the quality of heat customer maintenance and heat demand analysis.

[0059] Step S30 in the method provided in the embodiment of the present application includes:

[0060] According to the multiple heat use priorities, initial weight allocation is performed on N groups of heat demand analysis data in the heat demand analysis data set to obtain N first data weights, wherein the size of the heat use priority is positively correlated with the size of the first data weight;

[0061] According to the N first data weights, randomly select M groups of heating demand analysis data in the heating demand analysis data set as the first training data set to train the first heating demand analysis branch, where M is an integer greater than 1 and less than N;

[0062] Using the first training data set, testing the first heating demand parsing branch to obtain M first test results and a first parsing accuracy rate of the first heating demand parsing branch;

[0063] According to the M first test results and the first parsing accuracy, weight update calculation is performed on the M groups of first training data in the first training data set to obtain M initial second data weights;

[0064] Combining the M initial second data weights with NM first data weights of other heating demand analysis data in the heating demand analysis data set, performing weight normalization processing to obtain N second data weights;

[0065] Based on the N second data weights, continue to randomly select M groups of heating demand analysis data as the second training data set, and train the second heating demand analysis branch until multiple heating demand analysis branches are obtained through training;

[0066] The plurality of heat demand parsing branches and the plurality of branch parsing accuracies are integrated to obtain a heat demand parser.

[0067] In the embodiment of the present application, according to the heat priority of multiple heat users, initial weight allocation is performed on N groups of heat demand analysis data in the heat demand analysis data set to obtain N first data weights. N represents the total number of data in the heat demand analysis data set, which is an integer greater than 1. Each group of heat demand analysis data includes sample heat demand information and heating parameters. In the initial weight allocation, the size of the heat priority is positively correlated with the size of the first data weight.

[0068] Exemplarily, according to the multiple heat usage priorities of the multiple heat users, the heat demand analysis data corresponding to the multiple heat users are identified, each group of heat demand analysis data identifies a corresponding heat usage priority, and then the ratio of the heat usage priority of each group of heat demand analysis data to the sum of the N heat usage priorities identified by the N groups of heat demand analysis data is calculated as the first data weight of each group of heat demand analysis data, that is, the initial weight, to complete the initial weight allocation. The size of the heat usage priority is positively correlated with the size of the first data weight, and the sum of the first data weights of the N groups of heat demand analysis data is 1.

[0069] Furthermore, based on ensemble learning, multiple heating demand parsing branches are trained to form a final heating demand parser, and the first heating demand parsing branch is trained first.

[0070] Among them, according to the N first data weights, M groups of heat demand analysis data are randomly selected from the heat demand analysis data set to form the first training data set. M is an integer greater than 1 and less than N, representing the number of data selected in each training heat demand analysis branch. For example, M is 6000 and N is 10000. According to the N first data weights as the probability of N groups of heat demand analysis data being selected, and then according to the N selection probabilities, M groups of heat demand analysis data are randomly selected from the N groups of heat demand analysis data. When the heat priority of the heat-using object is large, the probability of the heat demand analysis data of the heat-using object being selected is large, and there are more heat demand analysis data of the heat-using object, then the first heat demand analysis branch trained is more accurate and efficient for the heat demand analysis of the heat-using object with greater heat adaptability. Random selection ensures the diversity of data and prevents the model from being overly dependent on certain specific data. At the same time, through weight guidance, it focuses on the sample data of heat-using objects with high heat priority.

[0071] The first training data set is used to train the first heating demand parsing branch. For example, the first heating demand parsing branch is first constructed based on the BP neural network. The first heating demand parsing branch is an untrained model and needs to be adjusted in parameters through training data to optimize its prediction ability.

[0072] The first training data set is input into the first heat demand parsing branch, and multiple rounds of training are performed. Each set of data in the first training data set (including heat demand characteristics and corresponding heating parameters) is input into the first heat demand parsing branch, and the predicted heat demand parsing result, i.e., the predicted heating parameter, is calculated according to the input data through the forward propagation process. The predicted result is compared with the actual heating parameter to calculate the prediction error or loss. The loss function (such as mean square error, cross entropy, etc.) is an indicator to measure the difference between the predicted result and the actual result. Through the back propagation algorithm, the error of the loss function is reversely transmitted, and the parameters of the first heat demand parsing branch are gradually adjusted to reduce the prediction error. After each round of training, the parameters of the first heat demand parsing branch are updated to better adapt to the training data. The above training steps are repeated many times, and each round of training will gradually improve the prediction ability of the first heat demand parsing branch until a preset stop condition is reached, such as reaching a preset number of training times, and the training of the first heat demand parsing branch is completed.

[0073] After the training is completed, the first training data set is used to test the first heating demand parsing branch to determine whether the heating parameters output by the first heating demand parsing branch are consistent with the actual sample heating parameters, and M first test results are obtained. Each first test result is correct or incorrect, 1 for correct and -1 for incorrect. The first parsing accuracy of the first heating demand parsing branch is obtained according to the proportion of the first test results that are correct in the M first test results. Optionally, the first parsing accuracy can also be obtained based on the ratio of the sum of the weights of the heating demand parsing corresponding to the first test result that is correct to the sum of the weights of all the data in the first training data set. Parsing accuracy is a key indicator to measure the performance of the first heating demand parsing branch on the test data, which reflects the model's ability to predict and parse heating demand.

[0074] In order to pay more attention to the heat demand parsing data that cannot be accurately parsed by the first heat demand parsing branch in the subsequent training process of other heat demand parsing branches, according to the M first test results and the first parsing accuracy, the weight update calculation is performed on the M groups of first training data in the first training data set to obtain M initial second data weights, and then the weights of the N groups of heat demand parsing data are normalized and updated to improve the performance of the overall heat demand parser.

[0075] The step in the method provided in the embodiment of the present application: "According to the M first test results and the first parsing accuracy, weight update calculation is performed on the M groups of first training data in the first training data set", which is calculated by the following formula:

[0076] ;

[0077] in, is the initial second data weight of the first training data of the i-th group, i=1,2,3…M, is the first data weight of the first training data of the i-th group, is the first analysis accuracy, is the first test result of the first training data of the i-th group. When the first test result is correct, is 1, the first test result is an error, is -1.

[0078] Through the above formula, according to the M first test results and the first parsing accuracy, the weight update calculation is performed on the M groups of first training data in the first training data set to obtain M initial second data weights. The weight update pays more attention to the heating demand parsing data that is difficult to accurately parse, increases its weight, and increases the probability of being selected in the subsequent training of other heating demand parsing branches, thereby improving the overall heating demand parser's parsing accuracy for these heating demand parsing data that are difficult to accurately parse. The weight of the data that has been accurately parsed is appropriately reduced to reduce the probability of being selected for training in the future, thereby guiding the heating demand parser to gradually improve its parsing ability for complex demands.

[0079] Further, the M initial second data weights and the NM first data weights of the NM groups of heating demand analysis data that are not selected in the other first training data sets in the heating demand analysis data set are combined to perform weight normalization processing to obtain N second data weights. The normalization processing includes making the sum of the M initial second data weights and the NM first data weights equal to 1, for example, calculating the ratio of each initial second data weight or first data weight to the sum of the M initial second data weights and the NM first data weights, obtaining N second data weights, and completing the processing.

[0080] Further, based on the same method, based on the N second data weights, M groups of heating demand parsing data are randomly selected from the heating demand parsing data set as the second training data set, and the second heating demand parsing branch is trained until multiple heating demand parsing branches are obtained through training. For example, the number of the multiple heating demand parsing branches is 20.

[0081] Finally, these multiple trained heat demand parsing branches and their respective parsing accuracies are integrated to build a more powerful heat demand parser. The training data of multiple heat demand parsing branches have certain intersections but do not overlap. During the training process, the weight of the selected data is set according to the heating priority, and the data weight is updated according to the verification results of multiple heat demand parsing branches. Serial training can gradually improve the parsing accuracy of the heat demand parsing branches, thereby improving the parsing accuracy of the overall heat demand parser and realizing complex demand parsing for multiple users. The heat demand parser can integrate the parsing results of multiple heat demand parsing branches to provide more accurate heat demand parsing capabilities, thereby ensuring that it can efficiently respond to various complex heat demand scenarios in actual applications.

[0082] S40: Based on the heating customer service platform, current heating demand information of the plurality of heating users is collected, and integrated analysis is performed based on the heating demand analyzer to obtain a plurality of heating parameters;

[0083] In the embodiment of the present application, in order to ensure that the needs of each heat-using object can be responded to in real time, the heat demand information of multiple heat-using objects is first collected based on the heat supply customer service platform.

[0084] After collecting these heat demand information, the heat demand parser is used for integrated analysis to obtain multiple heating parameters. The heat demand parser is a complex model that integrates multiple heat demand analysis branches. By integrating the analysis results of multiple heat demand analysis branches, it can more accurately predict and analyze the heat demand of each heat user and obtain accurate heating parameters as the basis for the actual operation of the heating system, ensuring that the system can respond to the needs of each heat user in real time and accurately, and improve the quality and efficiency of heating services.

[0085] Step S40 in the method provided in the embodiment of the present application includes:

[0086] Based on the heating customer service platform, collecting the current multiple heating demand information of the multiple heating users;

[0087] Inputting the plurality of heat demand information into the plurality of heat demand parsing branches in the heat demand parser, and parsing to obtain a plurality of parsed heating parameter sets;

[0088] According to the multiple branch analysis accuracy rates, the multiple analyzed heating parameter sets are weightedly fused to obtain multiple heating parameters.

[0089] In the embodiment of the present application, in order to accurately meet the needs of each heat user, the system first collects the current multiple heat demand information through the heat supply customer service platform. The multiple heat demand information includes the specific heat demand of each heat user at the current moment, such as high temperature or low temperature. As the center of data collection and user interaction, the heat supply customer service platform can collect these demand information in real time and organize it into structured data that can be processed by the system, such as text data.

[0090] Multiple heat demand information is input into multiple heat demand parsing branches in the heat demand parser, and heat demand parsing is performed respectively to obtain output results of multiple heat demand parsing branches, that is, multiple parsed heating parameter sets are obtained by parsing. Each parsed heating parameter set includes multiple parsed heating parameters obtained by parsing each heat demand information input into multiple heat demand parsing branches, such as specifically including heating temperature.

[0091] According to the multiple branch analysis accuracy, multiple analysis heating parameter sets are weighted and fused. The multiple branch analysis accuracy represents the accuracy of the heating demand analysis of multiple heating demand analysis branches. The ratio of each branch analysis accuracy to the sum of multiple branch analysis accuracy is calculated as a weight to obtain multiple branch weights. Then, according to the multiple branch weights, multiple analysis heating parameters in each analysis heating parameter set are weightedly calculated to obtain multiple heating parameters. For example, according to the multiple branch weights, the heating temperatures in the multiple analysis heating parameters in each analysis heating parameter set are weightedly calculated to obtain the weighted heating temperature as the analysis result of multiple heating demand information of multiple heat users.

[0092] By weighted fusion of multiple sets of analyzed heating parameters, a set of optimized heating parameters is obtained. These parameters will be used in actual heating scheduling to ensure that the needs of each heating object can be accurately met, while optimizing the utilization efficiency of heating resources. This process not only improves the response speed and quality of the heating demand analysis service, but also can cope with complex and changing heating scenarios.

[0093] S50: Based on the multiple heat usage priorities, multiple control orders are set, and the multiple heating parameters are impact analyzed and corrected to obtain multiple corrected heating parameters as heating demand analysis results, and customer feedback and heating control are performed.

[0094] In order to ensure that the heating service can accurately meet the needs of different heat users and give priority to the heating requirements of high-priority objects, multiple control orders will be set based on the multiple heat priorities obtained in the previous analysis. The control order refers to the execution order set for each heat parameter according to the priority of the heat user during the heating scheduling process.

[0095] After the control sequence is set, the system will perform impact analysis and correction on the generated multiple heating parameters. Impact analysis refers to analyzing the interactions and impacts that may occur in the actual execution of various heating parameters. For example, if a heating parameter affects the heating of adjacent heating objects, the heating parameters of adjacent heating objects can be adjusted by analyzing the impact, improving the overall heating accuracy of multiple heating objects and eliminating the heating impact of adjacent objects.

[0096] Through precise parameter control and priority management, the quality of heating services can be effectively improved, ensuring that the needs of high-priority objects are met first, while taking into account the heating needs of other objects, optimizing the utilization efficiency of heating resources, and reducing energy consumption and costs in the heating process.

[0097] Step S50 in the method provided in the embodiment of the present application includes:

[0098] Based on the plurality of heat usage priorities, setting a plurality of control orders in descending order;

[0099] According to the multiple control orders, the multiple heating parameters are used in turn to perform heating simulation, obtain the heating impact parameters on adjacent heating objects, adjust and correct the heating parameters of adjacent heating objects, obtain multiple corrected heating parameters, and perform customer feedback and heating control.

[0100] In the embodiment of the present application, in order to ensure that the needs of different heat-using objects are optimally met, especially objects with higher heat priority, multiple control orders are set in descending order based on multiple heat priorities, that is, multiple control orders are set in order from large to small, and the greater the heat priority, the higher the control order.

[0101] According to these orders, the corresponding heating parameters are used to perform heating simulation. Heating simulation is a virtual operation. Based on the digital heating simulation means in the prior art, the heating process can be simulated in a computer environment to pre-evaluate the impact of various heating parameters on actual heating. Through this simulation, the heating situation of each heating object can be predicted, and the impact of these heating operations on neighboring heating objects can be calculated. Neighboring heating objects refer to objects that are geographically or adjacent to the current heating object in the heating network, and their heating demand may be affected by high-priority objects. For example, when a large amount of heat is allocated to an object with a higher priority, it may have a thermal impact on its surrounding objects, causing the temperature to rise. After obtaining these heating impact parameters, the system will adjust and correct the heating parameters of the affected neighboring heating objects. Adjustment and correction means that the system eliminates or mitigates the negative impact of heating of higher-priority objects by adjusting the heating parameters of neighboring objects. For example, if the heating simulation shows that a neighboring object may have its temperature rise due to the high heating temperature of the high-priority object, the heating supply of the neighboring object is reduced or its heating time period is adjusted according to the extent of the temperature rise, and the heating parameters are adjusted to ensure that its heating demand is met. In this way, multiple corrected heating parameters are adjusted and used as the results of the heating demand analysis, and customer feedback and heating control are performed to complete the heating demand analysis and heating control. The extent of the temperature rise of the neighboring object due to the high heating temperature of the high-priority object, as well as the adjustment extent of the heating parameter correction, can be implemented based on the heating control in the prior art.

[0102] Among them, customer feedback refers to informing users of the results of analysis and adjustment, allowing them to understand how the heating system responds to their needs and further optimize services based on feedback. Heating control is to perform actual heating operations based on the finalized corrected heating parameters to ensure that the expected results can be achieved during operation, thereby improving the overall quality of heating services and user satisfaction.

[0103] Optionally, heating of multiple heat-consuming objects may be performed directly according to multiple control sequences and using multiple heating parameters.

[0104] The embodiments of the present application achieve the following technical effects through the above specific implementation methods:

[0105] The present invention collects object feature information of multiple heat-using objects in a heating system based on a heating customer service platform, performs heat priority analysis, obtains multiple heat priorities, and then collects historical heat usage data of multiple heat-using objects. Sample data is screened according to the multiple heat priorities to obtain a heating demand analysis data set, which can screen more heating demand analysis data of heat-using objects with larger priorities, thereby improving the accuracy of subsequent training of a heating demand analyzer for analyzing the heat demand of heat-using objects with larger priorities. Then, according to the multiple heat priorities, the heating demand analysis data set is used to train the heating demand analyzer, wherein initial data weights are set according to the multiple heat priorities, and the initial data weights are used during training. In the method, the data weight is adjusted according to the training results, the heating demand analysis data is randomly selected according to the weight, and the heating demand analyzer is trained by integrated learning, which can improve the overall accuracy of the heating demand analyzer. Based on the heating customer service platform, the current heating demand information of multiple heating objects is collected, and integrated analysis is performed based on the heating demand analyzer to obtain multiple heating parameters. Based on multiple heating priorities, multiple control orders are set, and the impact analysis and correction of multiple heating parameters are performed to obtain multiple corrected heating parameters as the heating demand analysis results, and customer feedback and heating control are performed to reduce the heating impact of neighboring heating objects on heating objects with higher priorities, thereby improving the quality of heating services. Through the above scheme, the present application considers the priorities of multiple heating objects, performs data collection and screening, and performs data weight setting and integrated training of the heating demand analyzer, which can improve the analysis accuracy and efficiency of the training heating demand analyzer, and the analysis accuracy of the demand with higher priority is greater, achieving the technical effect of improving the analysis accuracy and response efficiency of complex heating demands.

[0106] Example 2

[0107] like Figure 2 As shown, based on the same inventive concept as the method for implementing customer service of a heating system based on complex demand analysis provided in Example 1, an embodiment of the present invention provides a customer service implementation system for a heating system based on complex demand analysis, the system comprising:

[0108] The priority analysis module 11 is used to collect object feature information of multiple heat-using objects in the heating system based on the heating customer service platform, perform heat-using priority analysis, and obtain multiple heat-using priorities;

[0109] The data collection and screening module 12 is used to collect the historical heat consumption data of the plurality of heat consumption objects, and screen the sample data according to the plurality of heat consumption priorities to obtain a heat demand analysis data set, wherein the heat demand analysis data set includes a plurality of object demand analysis data sets;

[0110] A parsing training module 13 is used to train a heat demand parser using the heat demand parsing data set according to the multiple heat use priorities, wherein initial data weights are set according to the multiple heat use priorities, and the data weights are adjusted during training according to the training results;

[0111] The demand analysis module 14 is used to collect the current heat demand information of the plurality of heat users based on the heat supply customer service platform, and perform integrated analysis based on the heat supply demand analyzer to obtain a plurality of heat supply parameters;

[0112] The heating feedback module 15 is used to set multiple control orders based on the multiple heat usage priorities, perform impact analysis and correction on the multiple heating parameters, obtain multiple corrected heating parameters as heating demand analysis results, and perform customer feedback and heating control.

[0113] Furthermore, the priority analysis module 11 is also used for:

[0114] Based on the heating customer service platform, multiple object feature information of multiple heat-using objects in the heating system is collected;

[0115] According to the plurality of object feature information, a priority classification table is input to perform classification to obtain a plurality of heat usage priorities, wherein the priority classification table includes an index relationship between a plurality of sample object feature information and a plurality of sample priorities.

[0116] Furthermore, the data collection and screening module 12 is also used for:

[0117] According to the historical heat consumption data of the plurality of heat users stored in the heat supply customer service platform, a plurality of sample heat demand information sets are collected, and qualified heating parameters under different sample heat demand information are obtained to obtain a plurality of sample heating parameter sets;

[0118] Combining the plurality of sample heat demand information sets and the plurality of sample heating parameter sets to obtain a plurality of basic object demand analysis data sets;

[0119] According to the multiple heat usage priorities, a preset training data quantity N is allocated to obtain multiple object data quantities, the multiple basic object demand analysis data sets are randomly screened to obtain multiple object demand analysis data sets, and the heating demand analysis data set is integrated, where N is an integer greater than 1.

[0120] Furthermore, the analysis training module 13 is also used for:

[0121] According to the multiple heat use priorities, initial weight allocation is performed on N groups of heat demand analysis data in the heat demand analysis data set to obtain N first data weights, wherein the size of the heat use priority is positively correlated with the size of the first data weight;

[0122] According to the N first data weights, randomly select M groups of heating demand analysis data in the heating demand analysis data set as the first training data set to train the first heating demand analysis branch, where M is an integer greater than 1 and less than N;

[0123] Using the first training data set, testing the first heating demand parsing branch to obtain M first test results and a first parsing accuracy rate of the first heating demand parsing branch;

[0124] According to the M first test results and the first parsing accuracy, weight update calculation is performed on the M groups of first training data in the first training data set to obtain M initial second data weights;

[0125] Combining the M initial second data weights with NM first data weights of other heating demand analysis data in the heating demand analysis data set, performing weight normalization processing to obtain N second data weights;

[0126] Based on the N second data weights, continue to randomly select M groups of heating demand analysis data as the second training data set, and train the second heating demand analysis branch until multiple heating demand analysis branches are obtained through training;

[0127] The plurality of heat demand parsing branches and the plurality of branch parsing accuracies are integrated to obtain a heat demand parser.

[0128] According to the M first test results and the first parsing accuracy, weight update calculation is performed on the M groups of first training data in the first training data set to obtain M initial second data weights, as shown in the following formula:

[0129] ;

[0130] in, is the initial second data weight of the first training data of the i-th group, is the first data weight of the first training data of the i-th group, is the first analysis accuracy, is the first test result of the first training data of the i-th group. When the first test result is correct, is 1, the first test result is an error, is -1.

[0131] Furthermore, the demand analysis module 14 is also used for:

[0132] Based on the heating customer service platform, collecting the current multiple heating demand information of the multiple heating users;

[0133] Inputting the plurality of heat demand information into the plurality of heat demand parsing branches in the heat demand parser, and parsing to obtain a plurality of parsed heating parameter sets;

[0134] According to the multiple branch analysis accuracy rates, the multiple analyzed heating parameter sets are weightedly fused to obtain multiple heating parameters.

[0135] Furthermore, the heat supply feedback module 15 is also used for:

[0136] Based on the plurality of heat usage priorities, setting a plurality of control orders in descending order;

[0137] According to the multiple control orders, the multiple heating parameters are used in turn to perform heating simulation, obtain the heating impact parameters on adjacent heating objects, adjust and correct the heating parameters of adjacent heating objects, obtain multiple corrected heating parameters, and perform customer feedback and heating control.

[0138] See also Figure 3 , Figure 3 Schematic diagram of an electronic device provided by an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device 100, including a memory 110, a processor 120, and a computer program 111 stored in the memory 110 and executable on the processor 120. When the processor 120 executes the computer program 111, the steps of the method in Example 1 are implemented.

[0139] See also Figure 4 , Figure 4 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention. Figure 4 As shown, this embodiment provides a computer-readable storage medium 200 on which a computer program 211 is stored. When the computer program 211 is executed by a processor, the steps of the method in Embodiment 1 are implemented.

[0140] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0141] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0142] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0143] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0144] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0145] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0146] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for implementing customer service of a heating system based on complex demand analysis, characterized in that: The method comprises: Based on the heating customer service platform, the object feature information of multiple heat-using objects in the heating system is collected, and the heat-using priority analysis is performed to obtain multiple heat-using priorities; Collecting historical heat consumption data of the plurality of heat-consuming objects, and screening sample data according to the plurality of heat consumption priorities to obtain a heat demand analysis data set, wherein the heat demand analysis data set includes a plurality of object demand analysis data sets; According to the multiple heat usage priorities, the heat demand parsing data set is used to train a heat demand parser, wherein initial data weights are set according to the multiple heat usage priorities, and the data weights are adjusted during training according to the training results; Based on the heating customer service platform, current heating demand information of the plurality of heating users is collected, and integrated analysis is performed based on the heating demand analyzer to obtain a plurality of heating parameters; Based on the multiple heat use priorities, multiple control orders are set, and the multiple heating parameters are subjected to impact analysis and correction to obtain multiple corrected heating parameters as heating demand analysis results, and customer feedback and heating control are performed; Among them, based on the heating customer service platform, the object characteristic information of multiple heat-using objects in the heating system is collected to conduct heat priority analysis, including: Based on the heating customer service platform, multiple object feature information of multiple heat-using objects in the heating system is collected; According to the plurality of object feature information, a priority classification table is input to classify the plurality of heat use priorities, wherein the priority classification table includes an index relationship between a plurality of sample object feature information and a plurality of sample priorities; The historical heat consumption data of the plurality of heat consumption objects are collected, and sample data are screened according to the plurality of heat consumption priorities to obtain a heat demand analysis data set, including: According to the historical heat consumption data of the plurality of heat users stored in the heat supply customer service platform, a plurality of sample heat demand information sets are collected, and qualified heating parameters under different sample heat demand information are obtained to obtain a plurality of sample heating parameter sets; Combining the plurality of sample heat demand information sets and the plurality of sample heating parameter sets to obtain a plurality of basic object demand analysis data sets; According to the multiple heat use priorities, a preset number of training data N is allocated to obtain multiple numbers of object data, the multiple basic object demand analysis data sets are randomly screened to obtain multiple object demand analysis data sets, and the heat demand analysis data sets are integrated to obtain, where N is an integer greater than 1; According to the plurality of heat use priorities, the heat demand parsing data set is used to train a heat demand parser, including: According to the multiple heat use priorities, initial weight allocation is performed on N groups of heat demand analysis data in the heat demand analysis data set to obtain N first data weights, wherein the size of the heat use priority is positively correlated with the size of the first data weight; According to the N first data weights, randomly select M groups of heating demand analysis data in the heating demand analysis data set as the first training data set to train the first heating demand analysis branch, where M is an integer greater than 1 and less than N; Using the first training data set, testing the first heating demand parsing branch to obtain M first test results and a first parsing accuracy rate of the first heating demand parsing branch; According to the M first test results and the first parsing accuracy, weight update calculation is performed on the M groups of first training data in the first training data set to obtain M initial second data weights; Combining the M initial second data weights with NM first data weights of other heating demand analysis data in the heating demand analysis data set, performing weight normalization processing to obtain N second data weights; Based on the N second data weights, continue to randomly select M groups of heating demand analysis data as the second training data set, and train the second heating demand analysis branch until multiple heating demand analysis branches are obtained through training; Integrate the multiple heat demand parsing branches and multiple branch parsing accuracy rates to obtain a heat demand parser; According to the M first test results and the first parsing accuracy, weight update calculation is performed on the M groups of first training data in the first training data set to obtain M initial second data weights, as shown in the following formula: ； in, is the initial second data weight of the first training data of the i-th group, is the first data weight of the first training data of the i-th group, is the first analysis accuracy, is the first test result of the first training data of the i-th group. When the first test result is correct, is 1, the first test result is an error, is -1; Wherein, based on the heating customer service platform, the current heating demand information of the plurality of heating users is collected, and integrated analysis is performed based on the heating demand parser to obtain a plurality of heating parameters, including: Based on the heating customer service platform, collecting the current multiple heating demand information of the multiple heating users; Inputting the plurality of heat demand information into the plurality of heat demand parsing branches in the heat demand parser, and parsing to obtain a plurality of parsed heating parameter sets; According to the multiple branch analysis accuracy rates, weighted fusion is performed on the multiple analysis heating parameter sets to obtain multiple heating parameters; Wherein, based on the multiple heat use priorities, multiple control orders are set, the multiple heating parameters are subjected to impact analysis and correction, multiple corrected heating parameters are obtained as heating demand analysis results, and customer feedback and heating control are performed, including: Based on the plurality of heat usage priorities, setting a plurality of control orders in descending order; According to the multiple control orders, the multiple heating parameters are used in turn to perform heating simulation, obtain the heating impact parameters on adjacent heating objects, adjust and correct the heating parameters of adjacent heating objects, obtain multiple corrected heating parameters, and perform customer feedback and heating control.

2. A heating system customer service implementation system based on complex demand analysis, characterized in that: The system is used to implement the customer service implementation method of a heating system based on complex demand analysis as described in claim 1, and the system includes: The priority analysis module is used to collect object feature information of multiple heat-using objects in the heating system based on the heating customer service platform, perform heat-using priority analysis, and obtain multiple heat-using priorities; A data collection and screening module, used to collect historical heat consumption data of the plurality of heat consumption objects, and screen sample data according to the plurality of heat consumption priorities to obtain a heat demand analysis data set, wherein the heat demand analysis data set includes a plurality of object demand analysis data sets; A parsing training module, configured to train a heat demand parser using the heat demand parsing data set according to the plurality of heat use priorities, wherein initial data weights are set according to the plurality of heat use priorities, and the data weights are adjusted during training according to training results; A demand analysis module, for collecting the current heat demand information of the plurality of heat users based on the heat supply customer service platform, and performing integrated analysis based on the heat supply demand analyzer to obtain a plurality of heat supply parameters; The heating feedback module is used to set multiple control orders based on the multiple heat usage priorities, perform impact analysis and correction on the multiple heating parameters, obtain multiple corrected heating parameters as the heating demand analysis results, and perform customer feedback and heating control.

3. An electronic device, characterized in that: include: Memory for storing computer software programs; The processor is used to read and execute the computer software program, thereby implementing the customer service implementation method of the heating system based on complex demand analysis as described in claim 1.

4. A computer-readable storage medium, characterized in that: The storage medium stores a computer software program, which, when executed by a processor, implements the method for implementing customer service of a heating system based on complex demand analysis as claimed in claim 1.

Images