A smart city application scenario evaluation method based on multivariate linear regression

Abstract

The present invention discloses a smart city application scenario evaluation method based on multiple linear regression. The method constructs a set of quantitative evaluation systems for city attributes. Based on the multiple linear regression method, an evaluation method for smart city application scenarios is established, thereby calculating the fitness value of each smart city application scenario, solving the limitation of the single attribute in the existing smart city evaluation theory system on the overall development plan of the smart city. The method can improve the rationality of which application scenarios the smart city chooses for intelligent construction in practice, and provide good support for the construction of smart cities.

Description

Technical Field

[0001] The present invention proposes a smart city application scenario evaluation method based on multivariate linear regression, which belongs to the field of data analysis and smart city construction planning. Background Art

[0002] Urban attributes refer to the position and main functions of a city in the political, economic and social development of a certain region, country or even a larger scope. With the increasing level of productivity, the division of labor in society has become more and more detailed, and the functions of cities have also changed from simple to complex. Modern cities generally have multiple functions, but their roles are divided into large and small, and their functions are divided into primary and secondary. Only the main functions can have a decisive influence on the development of the city's nature and reflect the basic characteristics of the city.

[0003] At present, there are relatively few studies on the quantitative evaluation of urban attributes. For commercial attributes, Zhu Yanyao's "Quantitative Analysis of the Impact of the Protection of Large Sites in Xi'an on Urban Space" evaluates the commercial capacity of the surrounding areas of Xi'an from the perspective of urban road spatial structure; for ecological attributes, Liu Yaqin's "Research on the Ecological Attributes of Urban Residential Areas and Fuzzy Comprehensive Evaluation" established a residential ecological attribute evaluation system, and the structure of the evaluation system is an open three-level multi-factor comprehensive fuzzy evaluation model. The indicators of the evaluation system cover three aspects: the natural ecological attributes, economic ecological attributes, and social ecological attributes of residential areas. Sergey et al. decomposed the construction of smart cities into several projects and designed a project evaluation standard to rank the construction value of the projects. The standard also takes into account the strategic attributes of the city and the subjective needs of residents. Deng Xianfeng comprehensively considered the characteristics, construction paths and promotion methods of smart cities on the basis of the urban informatization evaluation theory, starting from the four major areas of industrial development, humanistic literacy, social services, and interconnection, and established secondary indicators in each field to build a highly recognized evaluation system. This system fully considers the vision and actual development status of smart cities based on successful theoretical experiences at home and abroad, and has a certain degree of scientificity.

[0004] The above scholars have actively explored the construction experience of smart cities from different levels and accumulated certain research results in the theoretical system of smart city evaluation, but they are more inclined to judge the single attribute of smart cities or the degree of intelligence, and have not made an overall evaluation of the application scenarios of smart cities.

[0005] The present invention proposes an effective solution to the above problems. The present invention constructs a quantitative evaluation system for urban attributes, and establishes an evaluation method for smart city application scenarios based on the multivariate linear regression method, thereby calculating the fitness value of each smart city application scenario, solving the limitation of the single attribute in the existing smart city evaluation theory system on the overall development plan of the smart city. Summary of the invention

[0006] In view of the defects existing in the prior art, the purpose of the present invention is to provide a smart city application scenario evaluation method based on multiple linear regression. The present invention designs a reasonable quantitative evaluation method to score highly abstract city attributes, and refers to the existing smart city sample data according to the actual situation of different application scenarios, and uses the multiple linear regression method to determine the weight value of each attribute, so as to linearly add each attribute to obtain the fitness value of different scenarios. This method not only performs reasonable quantitative scoring of city attributes, but also calculates the fitness value of different application scenarios.

[0007] The present invention analyzes the attributes of each application scenario of the smart city, calculates the fitness value of each application scenario according to the attribute scores under different application scenarios of the smart city, and finally, the decision maker can decide in which scenarios to carry out intelligent construction according to the size of the fitness value.

[0008] In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is: a smart city application scenario evaluation method based on multivariate linear regression.

[0009] The "smart city application scenario evaluation method based on multivariate linear regression" described in the present invention has the following steps:

[0010] Step A: Construct a set of smart city application scenarios;

[0011] Step B: Calculate the smart city feature matrix;

[0012] Step C: Calculate the regression coefficient of the multiple linear regression equation;

[0013] Step D: Calculate the fitness matrix of smart city application scenarios.

[0014] The specific meaning of "constructing a smart city application scenario set" in step A is to list the application scenarios of smart cities. The scenario set mainly includes the following scenarios:

[0015] (1) Smart firefighting: This scenario uses the latest technologies such as the Internet of Things, artificial intelligence, virtual reality, and mobile Internet+ to realize the intelligence of urban firefighting, improve the efficiency of information transmission, ensure the integrity of firefighting facilities, improve law enforcement and management effects, enhance rescue capabilities, and reduce the occurrence and losses of fires;

[0016] (2) Smart transportation: As a key and difficult aspect of urban management, the application of IoT technology improves the intelligence of traffic management and the operating efficiency of the entire urban transportation system. At the same time, it can fully demonstrate the comprehensive strength of the city.

[0017] (3) Smart factory: This scenario is a new stage in the development of modern factory informatization. Based on the digital factory, the Internet of Things technology and equipment monitoring technology are used to strengthen information management and services, and build a humanized factory that is energy-efficient, green, environmentally friendly, and comfortable.

[0018] (4) Smart security: In this scenario, big data processing technology and machine learning technology can be used to process and analyze the massive video, audio, and image data collected by the front end, greatly improving the accuracy and safety of security, and moving from passive defense to active judgment and early warning;

[0019] (5) Smart Tourism: This scenario uses new technologies such as cloud computing and the Internet of Things to actively perceive information on tourism resources, tourism economy, tourism activities, tourists, etc. through the Internet / mobile Internet and with the help of portable terminal Internet devices, and publishes it in a timely manner so that people can understand this information in a timely manner and arrange and adjust work and travel plans in a timely manner, thereby achieving the effect of intelligent perception and convenient use of various types of tourism information.

[0020] The specific meaning of "calculating the smart city feature matrix" in step B is: constructing an evaluation system for each attribute of a smart city; constructing a weight distribution system for city attributes in different scenarios; and calculating the feature matrix of different attributes in each scenario. The specific content of step B includes the following five steps:

[0021] Step B1: Calculate the characteristic matrix of urban strategic positioning under different scenarios;

[0022] Step B2: Calculate the characteristic matrix of typical industry development status;

[0023] Step B3: Calculate the characteristic matrix of urban functional zoning under different scenarios;

[0024] Step B4: Calculate the attribute matrix of urban resources;

[0025] Step B5: Calculate the attribute matrix of the city’s informationization degree.

[0026] Among them, the specific method of "calculating the attribute matrix of urban strategic positioning in different scenarios" described in step B1 is as follows: the impact of urban strategic positioning on different intelligent application scenarios is mainly measured by national policies, provincial policies and city policies. The scoring rules are as follows:

[0027] Table 1 Scoring rules for urban strategic positioning in different application scenarios

[0028]

[0029] For application scenario i, the score of city strategic positioning is equal to:

[0030]

[0031] Where j is the three different policy types in Table 1, x 1,j is the score of the jth policy. The evaluation method of urban strategic positioning is relatively simple. It only needs to count the policies at all levels related to a certain application scenario without weight allocation;

[0032] The specific method of "calculating the characteristic matrix of typical industry development status" in step B2 is as follows: for application scenario i, the score of typical industry development status (full score 100) is equal to:

[0033]

[0034] where w t is the weight of industry t, ranging from [0-1], x 2,t is the regional GDP share of industry t, and n is the number of typical industries. Since different application scenarios have different relevance to various industries (for example, smart logistics scenarios are more relevant to transportation, commerce, and industry, and less relevant to agriculture and tourism), the weight values ​​of each industry need to be determined when evaluating the development status of typical industries. In addition, considering that this attribute involves many different industries, the weight setting requires rich industry knowledge. Here, the expert experience method is used to directly give the corresponding weights for different industries;

[0035] The specific method of "calculating the characteristic matrix of urban functional zoning under different scenarios" in step B3 is as follows: If the application scenario i of the city has corresponding functional zoning, the functional zoning score is is 100, otherwise it is 0. The evaluation of urban functional zoning in different scenarios does not involve weight calculation. It only needs to determine whether there is a corresponding zoning for the application scenario, which will not be elaborated here;

[0036] The specific method of "calculating the characteristic matrix of urban resources" in step B4 is as follows: Urban resources can be divided into the following categories: historical culture, natural landscape, cultural landscape, educational resources, human resources, etc. Different resources have different impacts on different smart city application scenarios. For scenario i, the evaluation rules are as follows:

[0037]

[0038] where w m is the weight of resource m in scenario i, and its value range is [0-1], and the sum of the weights of all resources in scenario i is 1. 4,mis the score of resource m; the evaluation of urban resources involves multiple resources, and different resources will have more or less impact on the design of smart city application scenarios. It is necessary to combine the opinions of multiple industry experts to determine the resource weights in different scenarios. Therefore, the frequency statistics method is used here for calculation. The specific process is as follows:

[0039] Suppose resource set M = {m1, m2, m3, m4, m5}, and ask k relevant experts to independently propose the most appropriate weight for each resource in the resource set based on the weight allocation survey table (Table 3.3).

[0040] Table 3.3 Weight distribution survey table

[0041]

[0042] Where j is the expert number. According to the collected weight allocation survey table, a single factor weight statistical test is conducted on each resource. i In it weight w i,j (j=1,2,···,k) to find the maximum value w max and the minimum value w min . Choose an integer q and use the formula Calculate the group distance that divides the weights into q groups, and divide the weights into q groups from large to small, and calculate the frequency and frequency of the weights in each group. Finally, according to the frequency and frequency distribution, the median of the group with the largest frequency is generally taken as resource m. i The weight w i The integer q is set manually according to the experts who participate in the scoring. The more experts there are, the larger q is. When the number of experts is less than 100, the value of 8-15 can be used;

[0043] The specific method of "calculating the attribute matrix of the city's informatization degree" in step B5 is as follows: IDI is used to evaluate the informatization degree of a city. The informatization development index consists of 5 classification indexes and 12 specific indicators. The indicator system is as follows:

[0044] Table 2 Index system

[0045]

[0046] For scenario i, the calculation method of the informatization development index is:

[0047]

[0048] Where k is the number of classified indexes of information development, h is the number of indicators of the i-th index of information application level, and w is the number of indicators of the i-th index of information application level. i is the weight of the i-th index in the total index, P ijis the standardized value of the jth index of the i-th index. The weights of the five indexes need to be determined during the calculation. Referring to the China Information Yearbook published by the National Bureau of Statistics, the weights of the five indexes are 0.25, 0.25, 0.2, 0.2, and 0.1, respectively.

[0049] The specific meaning of "calculating the regression coefficient of the multivariate linear regression equation" in step C is: selecting relatively successful smart city application scenarios at home and abroad as sample data, and using the least squares method to fit the city attribute weights in each scenario. The specific content of step C includes the following five steps:

[0050] Step C1: Calculate the characteristic matrix of multiple successful smart cities;

[0051] Step C2: Experts judge the fitness value of each application scenario;

[0052] Step C3: Fitting the city attribute weights under each scenario;

[0053] The specific method of "calculating the characteristic matrices of multiple successful smart cities" in step C1 is as follows: select relatively successful smart city application scenarios at home and abroad as samples, execute steps A and B, and obtain the characteristic matrix X of the sample city. i ;

[0054] The specific method of "experts judging the fitness value of each application scenario" in step C2 is as follows: experts directly give the fitness matrix Y of each application scenario based on relevant data and experience. i ;

[0055] The specific method of "fitting the city attribute weights in each scenario" in step C3 is as follows: for the matrix obtained in steps C1 and C2, the linear regression coefficient is solved using the least squares method. The least squares method formula is as follows:

[0056]

[0057] where Y i It is the application scenario adaptation value of the observed sample. It is the application scenario adaptability of the fitting function. Taking the two-linear regression model as an example, the standard equation group for solving the regression parameters is:

[0058]

[0059] Solving this equation can obtain the values ​​of β0, β1, and β2. It can also be solved using the following matrix method:

[0060] β=(X T X) -1 X T Y

[0061] When using this method, scenarios with reasonable design and successful application should be selected as much as possible, which is of great significance to improving the evaluation quality of the evaluation model.

[0062] The "calculating the fitness matrix of smart city application scenarios" in step D means calculating the fitness value of smart city application scenarios. The specific method of step D is to multiply the city attribute feature matrix X obtained in step B by the weight matrix β in step C to obtain the application scenario fitness matrix Y.

[0063] The present invention has the following beneficial effects:

[0064] At present, there are few evaluation methods for smart city application scenarios. However, the intelligent construction of application scenarios is a crucial part of the development of smart cities. The present invention innovatively abstracts the fitness calculation problem of smart city application scenarios into a multivariate linear regression problem based on the actual constraints and development history of smart cities. Successful cases can be analyzed, and the city attribute weights can be more reasonably calculated through the least squares method, thereby improving the accuracy of the fitness of smart city application scenarios.

[0065] The present invention designs a complete urban attribute evaluation system when solving the problem. Different evaluation methods are designed according to different attributes to quantify urban attributes, which greatly improves the reliability of the fitness model and helps decision makers choose better application scenarios for intelligent construction.

[0066] The present invention adopts multiple linear regression when solving, which is simpler and more convenient when analyzing multi-factor models, and can efficiently and quickly obtain a solution set, and can calculate a unique result through standard statistical methods, ensuring that there will be no multiple solution sets in the result. At the same time, regression analysis can accurately measure the correlation between various factors and the degree of regression fit, thereby improving the credibility of the model.

[0067] In summary, this smart city application scenario evaluation method based on multivariate linear regression can improve the rationality of smart cities' selection of application scenarios for intelligent construction in practice, and provide good support for the construction of smart cities. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 Flowchart of smart city application scenario evaluation method DETAILED DESCRIPTION

[0069] In response to the current demand for evaluation methods for building smart city application scenarios, a multivariate linear regression application scenario evaluation method based on urban attributes is adopted. Figure 1This is a smart city application scenario evaluation method. The city attributes in this method include the strategic positioning of smart cities, typical industry development status, functional zoning, etc. These attributes are quantitatively scored using a variety of methods involved in the article. The application scenario set includes smart fire protection, smart transportation, etc. The fitness evaluation of the application scenario is achieved through the multivariate linear regression method. After the method is established, a reverse evaluation is performed by calculating the fitness of multiple smart application scenarios in the demonstration city to evaluate whether the city attribute quantification method and regression coefficient meet the scenario requirements.

[0070] In order to make the technical problems and technical solutions to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific implementation cases. It should be understood that the implementation examples described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0071] The purpose of the present invention is to provide a smart city application scenario evaluation method based on multiple linear regression. The present invention establishes an application scenario evaluation method based on the actual constraints and development history of actual smart cities, and on this basis uses a multiple linear regression algorithm to calculate weights, and then solves the application scenario fitness matrix. This method takes into account both the actual limitations of different attributes in smart cities and the speed and quality of problem solving, which can effectively improve the efficiency of intelligent construction of smart city application scenarios, reduce construction costs, and has good application value.

[0072] The present invention is further described below in conjunction with the accompanying drawings and specific implementation methods.

[0073] The implementation of the present invention takes the intelligent construction of application scenarios in Dali City and Yuxi City in Yunnan Province as an example to illustrate the method of the present invention. Dali City is an internationally renowned green tourism city with closely connected living areas and tourist areas, abundant renewable energy, green economy as the core industry, and is characterized by a beautiful environment, green livability and tourism. Yuxi City is the hometown of tobacco and an important industrial city in Yunnan. It has a complete industry, a complete industrial chain, many small and medium-sized enterprises, and a strong demand for energy. Faced with the problem of industrial pollution, Yuxi City has proposed a "high energy efficiency and low pollution" transformation strategy to achieve industrial energy transformation through "electricity substitution"; it is now necessary to analyze the urban attributes of these two cities and calculate the application scenario fitness matrix.

[0074] The present invention provides a smart city application scenario evaluation method based on multiple linear regression, such as Figure 1 As shown, the steps are as follows:

[0075] Step A: Construct a set of smart city application scenarios;

[0076] Step B: Calculate the smart city feature matrix;

[0077] Step C: Calculate the regression coefficient of the multiple linear regression equation;

[0078] Step D: Calculate the fitness matrix of smart city application scenarios.

[0079] The specific meaning of "constructing a smart city application scenario set" in step A is to list the application scenarios of smart cities. The scenario set mainly includes five scenarios: smart fire protection, smart transportation, smart factory, smart security, and smart tourism.

[0080] The specific meaning of "calculating the smart city feature matrix" in step B is: constructing an evaluation system for each attribute of a smart city; constructing a weight distribution system for city attributes in different scenarios; and calculating the feature matrix of different attributes in each scenario. The specific content of step B includes the following five steps:

[0081] Step B1: Calculate the characteristic matrix of urban strategic positioning under different scenarios;

[0082] Step B2: Calculate the characteristic matrix of typical industry development status;

[0083] Step B3: Calculate the characteristic matrix of urban functional zoning under different scenarios;

[0084] Step B4: Calculate the characteristic matrix of urban resources;

[0085] Step B5: Calculate the characteristic matrix of the city’s informatization degree.

[0086] The specific method of "calculating the characteristic matrix of the city's strategic positioning in different scenarios" in step B1 is as follows: score the two cities according to the scoring rules. For application scenario i, the score of the city's strategic positioning is equal to:

[0087]

[0088] Where j is the three different policy types in Table 1, x 1,j is the score of the jth policy. The evaluation method of urban strategic positioning is relatively simple. It only needs to count the policies at all levels related to a certain application scenario without weight allocation. After calculation, the attribute matrix of urban strategic positioning under different application scenarios of the two cities can be obtained as follows:

[0089]

[0090] The specific method of "calculating the characteristic matrix of typical industry development status" in step B2 is as follows: for application scenario i, the score of typical industry development status (full score 100) is equal to:

[0091]

[0092] where wt is the weight of industry t, ranging from [0-1], x 2,t is the regional GDP share of industry t, and n is the number of typical industries. Experts use empirical methods to directly give the corresponding weights of different industries. After calculation, the attribute matrix of the development status of typical industries in different application scenarios of the two cities can be obtained as follows:

[0093]

[0094] The specific method of "calculating the characteristic matrix of urban functional zoning under different scenarios" in step B3 is as follows: If the application scenario i of the city has corresponding functional zoning, the functional zoning score is The value is 100, otherwise it is 0. The evaluation of urban functional zoning in different scenarios does not involve weight calculation. It only needs to determine whether there is a corresponding zoning for the application scenario. The characteristic matrix of urban functional zoning in different application scenarios of the two cities is:

[0095]

[0096] The specific method of "calculating the characteristic matrix of urban resources" in step B4 is as follows: Urban resources can be divided into the following categories: historical culture, natural landscape, cultural landscape, educational resources, human resources, etc. Different resources have different impacts on different smart city application scenarios. For scenario i, the evaluation rules are as follows:

[0097]

[0098] where w m is the weight of resource m in scenario i, and its value range is [0-1], and the sum of the weights of all resources in scenario i is 1. 3,m is the score of resource m; the evaluation of urban resources involves multiple resources, and it is necessary to integrate the opinions of multiple industry experts to determine the resource weights in different scenarios. Therefore, the frequency statistics method is used here for calculation, and the specific process is not repeated here. The characteristic matrix of the two urban resources is:

[0099]

[0100] The specific method of "calculating the characteristic matrix of the city's informatization degree" in step B5 is as follows: IDI is used to evaluate the informatization degree of a city. The informatization development index consists of 5 classification indexes and 12 specific indicators. The indicator system is not repeated here. For scenario i, the calculation method of the informatization development index is:

[0101]

[0102] Where k is the number of classified indexes of information development, h is the number of indicators of the i-th index of information application level, and w is the number of indicators of the i-th index of information application level. i is the weight of the i-th index in the total index, P ij is the standardized value of the jth index of the i-th index. In addition, the weights of the five indexes refer to the "China Information Yearbook" published by the National Bureau of Statistics, which are 0.25, 0.25, 0.2, 0.2, and 0.1 respectively. After calculation, the characteristic matrix of the informatization degree of the two cities can be obtained as follows:

[0103]

[0104] The specific meaning of "calculating the regression coefficient of the multivariate linear regression equation" in step C is: selecting relatively successful smart city application scenarios at home and abroad as sample data, and using the least squares method to fit the city attribute weights in each scenario. The specific content of step C includes the following five steps:

[0105] Step C1: Calculate the characteristic matrix of multiple successful smart cities;

[0106] Step C2: Experts judge the fitness value of each application scenario;

[0107] Step C3: Fitting the city attribute weights under each scenario;

[0108] The specific method of "calculating the characteristic matrices of multiple successful smart cities" in step C1 is as follows: select relatively successful smart city application scenarios at home and abroad as samples, execute steps A and B, and obtain the characteristic matrix X of the sample city. i ;

[0109] The specific method of "experts judging the fitness value of each application scenario" in step C2 is as follows: experts directly give the fitness matrix Y of each application scenario based on relevant data and experience. i ;

[0110] The specific method of "fitting the city attribute weights in each scenario" in step C3 is as follows: for the matrix obtained in steps C1 and C2, the linear regression coefficient is solved using the least squares method. The least squares method formula is as follows:

[0111]

[0112] where Y i It is the application scenario adaptation value of the observed sample. It is the application scenario adaptability of the fitting function. Taking the two-linear regression model as an example, the standard equation group for solving the regression parameters is:

[0113]

[0114] Solving this equation can obtain the values ​​of β0, β1, and β2. It can also be solved using the following matrix method:

[0115] β=(X T X) -1 X T Y

[0116] The weight matrix obtained after calculation is:

[0117]

[0118] The "calculation of the fitness matrix of smart city application scenarios" in step D means: calculating the fitness value of smart city application scenarios. The specific method of step D is: multiplying the city attribute feature matrix obtained in step B with the weight matrix in step C to obtain the application scenario fitness matrix:

[0119]

[0120] Each row of the fitness matrix represents an application scenario, which are smart fire protection, smart transportation, smart factory, smart security and smart tourism.

[0121] From the application scenario fitness matrix, we can see that Dali is most suitable for developing smart tourism, followed by smart fire protection. Since Dali has rich tourism resources such as Cangshan Mountain, Erhai Lake, Shuanglang, and a developed tourism industry, and the city's strategic positioning is a tourist city and a livable city, it is reasonable that Dali's smart tourism and fire protection application scenarios have the highest scores. As Yuxi is the hometown of tobacco, the tobacco industry is developed, and it is rich in natural resources. The city's strategic positioning is a garden city and an industrial city, so it is reasonable to build smart factories and smart transportation in Yuxi.

[0122] The parts not described in detail in the present invention belong to the well-known technology in the art.

[0123] The above descriptions are only some specific implementation methods of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any person familiar with the art within the technical scope disclosed in the present invention should be covered within the protection scope of the present invention.

Claims

1. A smart city application scenario evaluation method based on multivariate linear regression, the steps are as follows: Step A: Construct a set of smart city application scenarios; Step B: Calculate the smart city feature matrix; Step C: Calculate the regression coefficient of the multiple linear regression equation; Step D: Calculate the fitness matrix of smart city application scenarios; The specific meaning of the step B is: constructing an evaluation system for each attribute of a smart city; constructing a weight distribution system for city attributes in different scenarios; and calculating a feature matrix of different attributes in each scenario; The specific content includes the following five steps: Step B1: Calculate the characteristic matrix of urban strategic positioning under different scenarios; Step B2: Calculate the characteristic matrix of typical industry development status; Step B3: Calculate the characteristic matrix of urban functional zoning under different scenarios; Step B4: Calculate the attribute matrix of urban resources; Step B5: Calculate the attribute matrix of the city's informatization degree; The specific steps of step B1 are as follows: The impact of urban strategic positioning on different smart application scenarios is mainly measured by three types of policies: national policy, provincial policy and city policy; For application scenario i, city strategic positioning The score calculation formula is as follows: Among them, j is 3 different policy types, x 1,j is the score of the j-th policy; The specific steps of step B2 are as follows: For application scenario i, the score calculation formula for the typical industry development status is as follows: where w t is the weight of industry t, ranging from [0-1], x 2,t is the share of regional GDP in industry t, and n is the number of typical industries; The specific steps of step B3 are as follows: If the application scenario i of the city has corresponding functional zones, the functional zone score is is 100, otherwise it is 0; The specific steps of step B4 are as follows: Urban resources are refined into: historical culture, natural landscape, cultural landscape, educational resources, and human resources. For scene i, the evaluation rules are as follows: where w m is the weight of resource m in scenario i, and its value range is [0-1], and the sum of the weights of all resources in scenario i is 1. 4,m is the score of resource m; The specific steps of step B5 are as follows: IDI is used to evaluate the informatization level of a city. For scenario i, the calculation formula of the informatization development index is: Where k is the number of classified indexes of information development, h is the number of indicators of the i-th index of information application level, and w is the number of indicators of the i-th index of information application level. i is the weight of the i-th index in the total index, P ij is the normalized value of the jth index of the i-th category index; The specific meaning of step C is: select relatively successful smart city application scenarios at home and abroad as sample data, and use the least squares method to fit the city attribute weights under each scenario. The specific content of step C includes the following five steps: Step C1: Calculate the characteristic matrix of multiple successful smart cities; Step C2: Experts judge the fitness value of each application scenario; Step C3: Fitting the city attribute weights under each scenario; The specific method of step C1 is: select relatively successful smart city application scenarios at home and abroad as samples, execute steps A and B, and obtain the characteristic matrix X of the sample city i ; The specific steps of step C2 are as follows: Experts directly provide the fitness matrix Y of each application scenario based on relevant data and experience i ; The specific method of step C3 is as follows: for the matrices obtained in step C1 and step C2, the linear regression coefficient is solved using the least square method. The least square method formula is as follows: where Y i It is the application scenario adaptation value of the observed sample. It is the application scenario adaptability value of the fitting function; Taking the two linear regression models as an example, the standard set of equations for solving the regression parameters is: Solve this equation to obtain the values ​​of β0, β1, and β2, or use the following matrix method to solve: β=(X T X) -1 X T Y; The meaning of step D is: calculating the fitness value of the smart city application scenario. The specific method is: multiplying the city attribute feature matrix X obtained in step B with the weight matrix β in step C to obtain the application scenario fitness matrix Y.

2. According to claim 1, a smart city application scenario evaluation method based on multivariate linear regression is characterized in that The step B4 uses the frequency statistics method to perform calculations, and the specific process is as follows: Assume that the resource set M = {m1, m2, m3, m4, m5}, and ask k experts to independently propose weights for each resource in the resource set according to the weight allocation survey table, where j is the expert number. According to the collected weight allocation survey table, a single-factor weight statistical test is conducted on each resource. First, resource m i In it weight w i,j (j=1,2,…,k) to find the maximum value w max and the minimum value w min , choose an integer q, and use the formula Calculate the group distance that divides the weights into q groups, and divide the weights into q groups from large to small, and calculate the frequency and frequency of the weights in each group. Finally, according to the frequency and frequency distribution, take the median value of the group with the maximum frequency as resource m i The weight w i The integer q is artificially set according to the experts participating in the scoring. The more experts there are, the larger q is. When the number of experts is less than 100, the value is 8-15.

Images