A cold chain transportation multi-source operation data processing method and system

Abstract

The application discloses a cold chain transportation multi-source operation data processing method and system, and belongs to the technical field of cold chain logistics operation data processing methods, which comprises the following steps: acquiring economic and social environment data, industry development basic data, vehicle operation data, temperature and humidity sensor data and transportation OD data of a target region in a target period, performing cleaning, correlation, index calculation and dimensionless processing to form an evaluation index set; constructing a complex weighted network based on the transportation OD data and calculating network development indexes; determining combined weights by using a combination of analytic hierarchy process, CRITIC method and game theory; generating a comprehensive evaluation cloud according to the combined weights and the cloud digital characteristics of each index, and performing similarity matching with a preset grade standard cloud to obtain a cold chain transportation operation evaluation grade; and calculating the obstacle degree of each level according to the index deviation degree and the combined weights, and generating an obstacle factor sorting result; the method can uniformly process multi-source heterogeneous cold chain data, and improve the comparability and traceability of operation evaluation results.

Description

Technical Field

[0001] This application belongs to the field of cold chain logistics operation data processing technology, specifically a method and system for processing multi-source operation data of cold chain transportation. Background Technology

[0002] During cold chain transportation, vehicle-mounted terminals, temperature and humidity sensors, and road cold chain transportation platforms continuously generate operational data such as vehicle operating status, cumulative mileage, runtime, cargo compartment temperature and humidity, over-temperature alarms, and transportation origin and destination. When processing the operational status of cold chain transportation within a specific area, it is also necessary to read economic and social statistical data, transportation industry statistical data, cold storage facility data, and refrigerated vehicle data. These data originate from statistical information systems, industry management systems, road cold chain transportation platforms, vehicle-mounted terminals, and temperature and humidity sensors, and their field identifiers, units of measurement, generation cycles, sampling frequencies, and spatial granularity are inconsistent.

[0003] Current cold chain transportation data processing methods typically involve statistical analysis based on data sources. For example, economic and social data and industry-specific basic data are aggregated annually or quarterly; vehicle operation data is aggregated by vehicle or calendar day; temperature and humidity data is calculated based on sensor sampling records to determine temperature control compliance; and transportation origin-destination data is compiled based on orders or transportation tasks to determine transportation volume and number of transport vehicles. Because there is a lack of unified vehicle identification, transportation task identification, and time aggregation benchmarks across different data sets, vehicle operation records, temperature and humidity sampling records, and transportation origin-destination records for the same transportation task may be assigned to different statistical periods. Duplicate records, missing records, or inconsistent timestamps can also lead to discrepancies in runtime, temperature control compliance duration, and the number of over-temperature alarms.

[0004] Transportation origin-destination data is typically stored in the form of individual orders, vehicle tasks, or operational trajectories. While it reflects the origin and destination of a transportation activity, it doesn't directly reflect the overall connectivity between different transportation nodes. Existing technologies that only aggregate total transportation volume, number of transport vehicles, number of refrigerated vehicles, or cold storage capacity lose the correspondence between transportation origins and destinations. For example, two regions may have the same total transportation volume and number of transport vehicles within a target period, but one region's transportation activity might be concentrated on a few nodes and rely on a single transit node, while the other region's activity might be distributed across multiple interconnected nodes. Relying solely on total transportation volume makes it difficult to distinguish these different transportation network states and to identify the potential impact of local node failures on regional transportation connectivity.

[0005] Furthermore, economic and social data, industry-specific data, vehicle operation data, temperature control data, and transportation network data each possess different numerical ranges, indicator directions, and degrees of dispersion. Some data may also repeatedly represent similar operational states. Directly splicing or accumulating various types of data can easily lead to duplicate effects in the processing results, especially for data with large numerical ranges or high interrelationships. Outputting different statistical results separately makes it difficult to determine whether data from different sources correspond to the same target period and the same transportation operation state. When processing results change, it is also difficult to trace whether the change is caused by vehicle operation status, temperature control status, transportation node connectivity, or data version changes, resulting in a lack of a consistent basis for comparison between processing results from different regions or periods.

[0006] Therefore, the technical problem that the existing technology needs to solve is: how to uniformly process cold chain transportation data with different sources, different time granularities and different data structures while preserving the correlation between vehicles, transportation tasks, collection time and transportation origin and destination, and convert discrete transportation origin and destination records into data that can characterize the spatial connection status of transportation, so that the resulting cold chain transportation operation processing results are comparable and the data source is traceable. Summary of the Invention

[0007] Given that existing cold chain transportation operation data are generated from statistical information systems, industry management systems, road cold chain transportation platforms, vehicle terminals, and temperature and humidity sensors, and that these different data vary in update cycles, spatial granularity, units of measurement, and data structures, and that transportation OD data cannot directly reflect the spatial connections and network functions between transportation nodes, this application provides a method and system for processing multi-source cold chain transportation operation data. By uniformly processing multi-source heterogeneous cold chain data, an evaluation index system is constructed that includes regional basic conditions, industry development status, actual vehicle operating status, and transportation network structure. Within the same processing chain, index weighting, operational level determination, and obstacle factor identification are completed to generate data processing results that simultaneously reflect the overall operational level and major limiting factors.

[0008] To solve the above-mentioned technical problems, this application adopts the following technical solution.

[0009] Firstly, this application provides a method for processing multi-source operation data in cold chain transportation, including:

[0010] Acquire multi-source heterogeneous cold chain data formed in the target area within the target period. The multi-source heterogeneous cold chain data includes data reflecting the regional economic and social conditions, the development foundation of the cold chain industry, the operating status of refrigerated vehicles, the temperature and humidity status of the cargo compartment, and the transportation origin and destination connections.

[0011] The multi-source heterogeneous cold chain data is cleaned, forward consistent, dimensionless, and indexed to form an evaluation index set organized according to the criteria layer, element layer, and index layer. The evaluation index set includes economic and social environment indicators, industry development indicators, industry operation indicators, and network development indicators.

[0012] The subjective weights of each indicator in the evaluation index set are determined by the analytic hierarchy process (AHP), the objective weights of each indicator are determined by the CRITIC method, and the subjective and objective weights are fused by game theory combination weighting to obtain the combined weights of each indicator.

[0013] A comprehensive evaluation cloud is generated based on the combined weights and cloud digital features corresponding to each indicator. The similarity between the comprehensive evaluation cloud and multiple preset level standard clouds is determined, and the evaluation level corresponding to the preset level standard cloud whose similarity meets the preset level determination conditions is determined as the cold chain transportation operation evaluation level.

[0014] The index deviation is determined based on the index values ​​after positive uniformity and dimensionless transformation of each index. Then, the index layer obstacle degree, element layer obstacle degree, and criterion layer obstacle degree are calculated in sequence based on the index deviation degree and the combined weight. The obstacle factor ranking results are generated according to the obstacle degree of each level.

[0015] The cold chain transportation operation evaluation level and the ranking results of the obstacle factors are written into the cold chain transportation multi-source operation data processing results and output.

[0016] In one implementation, when acquiring the multi-source heterogeneous cold chain data, the corresponding statistical versions of economic, social and environmental data and industry development basic data are read according to the target cycle, and vehicle operation data, temperature and humidity sensor data and transportation OD data are collected according to the data generation time; the association between different data records is established based on vehicle identification, transportation task identification and collection time, so that data from different sources and with different collection frequencies can be attributed to the same target area, target cycle or transportation task.

[0017] In one implementation, the multi-source heterogeneous cold chain data is cleaned, including standardizing field names and formats, converting units of measurement, merging or deleting duplicate records, filling in missing records or marking them as invalid, and identifying invalid records based on vehicle identification, transportation task identification, collection time, and data value range. After calculating indicators on the cleaned data, positive consistency processing is performed according to the indicator direction, and each indicator value is converted to a unified evaluation domain.

[0018] In one implementation, the industry operation indicators include at least one of vehicle operation rate, number of transport vehicles, average daily vehicle mileage, average daily vehicle operating time, overall temperature control compliance rate, vehicle over-temperature alarm rate, and number of vehicle over-temperature alarms.

[0019] In one implementation, the network development indicators are determined based on transportation origin-destination (OD) data. Specifically, transportation origins, destinations, or node regions are designated as network nodes, and transportation connections between different network nodes are designated as network edges. Network edge weights are set based on transportation volume, frequency, or number of transport vehicles. Depending on whether the transportation direction participates in network calculation, a directed weighted network or an undirected weighted network is constructed. Based on the constructed complex weighted network, at least one of the following network development indicators is calculated: spatial concentration, network density, network clustering, network connectivity efficiency, node connectivity, network intermediation, and connectivity convenience.

[0020] In one implementation, when using the analytic hierarchy process (AHP) to determine the subjective weights, a criterion-level judgment matrix, an element-level judgment matrix, and an indicator-level judgment matrix are constructed respectively. Pairwise comparisons are performed on the evaluation objects at the same level according to a preset importance scale. A consistency test is performed on each judgment matrix. When the consistency ratio is less than a preset consistency threshold, the local subjective weights at different levels are synthesized step by step to form the global subjective weights of each indicator.

[0021] In one implementation, when using the CRITIC method to determine the objective weights, multiple regional evaluation samples or multiple periodic evaluation samples are read, and the standard deviation of each indicator in the evaluation samples and the correlation coefficient between different indicators are calculated. The standard deviation characterizes the difference in the values ​​of the indicators, and the correlation coefficient characterizes the degree of information overlap between the indicators. Based on this, the information content of each indicator is determined, and the corresponding objective weights are obtained after normalizing the information content.

[0022] In one implementation, when determining the combined weights using game theory-based combinational weighting, the global subjective weights and the objective weights are used as basic weight vectors. The goal is to achieve equilibrium between the deviations between the combined weight vectors and the basic weight vectors. The combination coefficients corresponding to each basic weight vector are then solved. The combination coefficients are normalized, and the normalized combination coefficients are used to perform a linear combination of the basic weight vectors to obtain the combined weights.

[0023] In one implementation, the evaluation domain is pre-divided into multiple evaluation levels, and the expected value and entropy of a preset level standard cloud are determined based on the upper and lower boundaries of the corresponding intervals for each evaluation level. The hyperentropy of the preset level standard cloud can be determined based on historical evaluation samples for the corresponding level; when historical evaluation samples are lacking for the corresponding level, it can be determined based on a preset hyperentropy coefficient and the entropy. After standardizing the cloud digital features of the comprehensive evaluation cloud and each preset level standard cloud, the Euclidean distance between the comprehensive evaluation cloud and each preset level standard cloud is calculated, and the cloud similarity is determined based on the Euclidean distance. The evaluation level corresponding to the maximum cloud similarity is determined as the cold chain transportation operation evaluation level.

[0024] In one implementation, the deviation of each indicator from the target state is determined based on the indicator values ​​after positive uniformity and dimensionless transformation. The indicator deviation of each indicator is combined with the corresponding combined weight to obtain the indicator layer obstacle degree. The indicator layer obstacle degrees belonging to the same element layer are summarized to obtain the element layer obstacle degree. The element layer obstacle degrees belonging to the same criterion layer are summarized to obtain the criterion layer obstacle degree. The obstacle degrees of the indicator layer, element layer, and criterion layer are sorted from largest to smallest to form the obstacle factor ranking result.

[0025] Secondly, this application provides a multi-source operation data processing system for cold chain transportation, including a data acquisition module, an indicator processing module, a combined weighting module, a cloud model evaluation module, an obstacle identification module, and a result output module.

[0026] The data acquisition module is used to acquire multi-source heterogeneous cold chain data of the target area within the target period and establish the correlation between different data records.

[0027] The indicator processing module is used to clean, positively consistent, dimensionless, and calculate indicators for the multi-source heterogeneous cold chain data, resulting in an evaluation indicator set including economic and social environmental indicators, industry development indicators, industry operation indicators, and network development indicators.

[0028] The combined weighting module is used to determine subjective weights using the analytic hierarchy process (AHP), determine objective weights using the CRITIC method, and combine the subjective and objective weights using game theory to obtain combined weights.

[0029] The cloud model evaluation module is used to generate a comprehensive evaluation cloud based on the combined weights and cloud digital features of each indicator, and to determine the cold chain transportation operation evaluation level based on the similarity between the comprehensive evaluation cloud and the preset level standard cloud.

[0030] The obstacle identification module is used to determine the index deviation degree based on the dimensionless index value, calculate the obstacle degree of the index layer, element layer and criterion layer based on the index deviation degree and the combined weight, and generate the obstacle factor ranking result.

[0031] The result output module is used to output the multi-source operation data processing results of cold chain transportation, including the cold chain transportation operation evaluation level and the obstacle factor ranking results.

[0032] Compared with related technologies, this application has the following advantages:

[0033] This application links economic and social environmental data, industry development data, vehicle operation data, temperature and humidity sensor data, and transportation OD data based on vehicle identification, transportation task identification, and collection time. The linked data undergoes cleaning, index calculation, positive consistency, and dimensionless processing. This transforms data from different statistical periods, sampling frequencies, and units of measurement into a set of evaluation indicators with a unified direction and numerical range, reducing the impact of data mismatch, duplicate records, and dimensional differences on the operational status processing results. By defining transportation origins and destinations as network nodes, and transportation connections between nodes as network edges, and determining the weight of network edges based on transportation volume, frequency, or number of transport vehicles, the spatial connectivity in the original transportation OD records can be transformed into spatial concentration, network density, network clustering, network connectivity efficiency, and node function indicators. This allows for the differentiation of data with similar total transportation volumes but... The cold chain transportation network exhibits varying degrees of node concentration, network connectivity, and the role of key nodes. By employing the Analytic Hierarchy Process (AHP) and the CRITIC method, basic weights reflecting the hierarchical relationships of indicators and the degree of sample differences are formed. Game theory-based weighting is then used to generate combined weights, ensuring that a single subjective judgment or the degree of dispersion of a single data point does not directly determine the processing result. Furthermore, a comprehensive evaluation cloud is generated based on these combined weights and matched with a preset level standard cloud for similarity analysis. Simultaneously, the same combined weights and dimensionless indicator values ​​are used to calculate the obstacle levels at the indicator, element, and criterion levels. This ensures that the cold chain transportation operation evaluation level and the obstacle factor ranking results have a consistent data foundation and calculation basis. Consequently, the output results not only characterize the overall operational status of the target area within the target period but also trace back to specific operational indicators, temperature control status, or transportation network structure, reducing the risk of outputting only a single level without explaining the source of data changes. Attached Figure Description

[0034] Figure 1 A flowchart illustrating a multi-source operation data processing method for cold chain transportation provided in this application embodiment;

[0035] Figure 2 A schematic diagram of a cold chain transportation evaluation index system structure provided in this application embodiment;

[0036] Figure 3 This application provides a schematic diagram of a networked processing flow for transportation OD data.

[0037] Figure 4 This application provides a schematic diagram of a weighting process for a combination of evaluation indicators.

[0038] Figure 5 This application provides a schematic diagram of a cloud model evaluation and obstacle factor identification process.

[0039] Figure 6 A schematic diagram of the structure of a multi-source operation data processing system for cold chain transportation provided in this application embodiment;

[0040] Figure 7 This is a schematic diagram illustrating an application scenario of a multi-source operation data processing system for cold chain transportation, provided in an embodiment of this application. Detailed Implementation

[0041] The technical solution of this application will be clearly and completely described below with reference to the accompanying drawings. The described embodiments are for illustrative purposes only and are not intended to limit the scope of protection of this application. Where there is no conflict, technical features in different embodiments can be combined with each other.

[0042] The target area referred to in this application may be a provincial-level administrative region, a prefecture-level administrative region, a county-level administrative region, a cold chain logistics hub coverage area, or an area divided according to the scope of cold chain transportation business.

[0043] The target period referred to in this application may be an annual, quarterly, monthly, weekly, or statistical period set according to the cold chain transportation operation monitoring task.

[0044] The transportation OD data referred to in this application refers to data used to characterize the origin, destination, and transportation links between cold chain transportation activities. OD stands for Origin and Destination, respectively.

[0045] The term "preset" as used in this application refers to parameters, conditions, or rules that have been determined before the current processing step is executed. The basis for their formation includes industry standards, historical samples, platform operation rules, sensor ranges, temperature control requirements for transported goods, or confirmed evaluation task configurations.

[0046] Data acquisition, field transformation, Min-Max dimensionless conversion, shortest path calculation, analytic hierarchy process, CRITIC method, and cloud generator can all be implemented using existing data processing methods.

[0047] See Figure 7This application provides a multi-source cold chain transportation operation data processing system that can be applied to cold chain transportation operation data processing scenarios in target areas. A statistical information system provides economic, social, and environmental data to the system; an industry statistical system provides basic industry development data; a road cold chain transportation platform provides transportation origin-destination (OD) data; an on-board terminal provides vehicle operation data; and temperature and humidity sensors installed in the cargo compartment of refrigerated vehicles provide temperature and humidity sensor data.

[0048] Within the target area, refrigerated vehicles perform cold chain transportation tasks along transportation routes between the transportation origin, cold chain logistics nodes or cold storage facilities, and the transportation destination. The road cold chain transportation platform generates transportation origin-destination (OD) data based on transportation order records, vehicle transportation task records, or vehicle travel trajectories; the on-board terminal collects vehicle identification, collection time, vehicle operating status, cumulative mileage, and runtime during the operation of the refrigerated vehicles; and the temperature and humidity sensors collect temperature and humidity values ​​inside the vehicle's cargo compartment according to the corresponding sampling period.

[0049] The cold chain transportation multi-source operation data processing system reads the above-mentioned multi-source heterogeneous cold chain data, and sequentially performs data acquisition, indicator processing, network construction, combined weighting, cloud model evaluation, and obstacle identification to generate cold chain transportation operation evaluation level and obstacle factor ranking results, and sends the cold chain transportation operation evaluation level and obstacle factor ranking results to the management terminal for display. Figure 7 The application scenario shown illustrates the data flow between various data sources, cold chain transportation activities within the target area, data processing systems, and management terminals.

[0050] Example 1

[0051] See Figure 1 The cold chain transportation multi-source operation data processing method provided in this embodiment includes steps S101 to S106.

[0052] S101: Obtain multi-source heterogeneous cold chain data for the target area within the target period.

[0053] The multi-source heterogeneous cold chain data includes economic, social, and environmental data, industry development data, vehicle operation data, temperature and humidity sensor data, and transportation OD data.

[0054] Economic, social and environmental data mainly come from statistical yearbooks, statistical bulletins or macroeconomic statistical information systems, including per capita GDP, number of permanent residents, proportion of the tertiary industry, total output value of agriculture, forestry, animal husbandry and fishery, per capita disposable income, total retail sales of consumer goods, number of mobile phone users and number of broadband Internet access.

[0055] The basic data for industry development mainly come from the transportation industry statistics system, the cold chain logistics industry statistics system, or the road cold chain transportation platform, including highway mileage, freight volume, cold storage capacity, number of refrigerated trucks, number of cold chain logistics enterprises, number of transportation practitioners, and the proportion of new energy refrigerated trucks.

[0056] Vehicle operation data is acquired by onboard terminals installed on refrigerated vehicles, satellite positioning terminals, or road cold chain transportation platforms. The vehicle operation data includes at least vehicle identification, collection time, vehicle operating status, cumulative mileage, and runtime.

[0057] Temperature and humidity sensor data is acquired by temperature and humidity sensors installed inside the cargo compartment of refrigerated vehicles. The temperature and humidity sensor data includes at least vehicle identification, acquisition time, temperature value, humidity value, and alarm indicator. The temperature value is used to determine the duration of temperature control compliance and over-temperature alarm events; the humidity value, as part of the corresponding sensor record, is used together with the vehicle identification, acquisition time, and temperature value for verifying the integrity of the sensor record.

[0058] Transportation origin-destination (OD) data is generated from transportation order records, vehicle transportation task records, or vehicle operation trajectories, and includes at least the transportation origin, transportation destination, transportation time, transportation volume, transportation frequency, and transportation vehicle identification.

[0059] The system establishes a correlation between vehicle operation data, temperature and humidity sensor data, and transportation origin-destination (OD) data based on vehicle identification, transportation task identification, and data collection time. Vehicle identification determines whether different data records correspond to the same vehicle, transportation task identification determines whether different data records correspond to the same transportation task, and data collection time is used to aggregate vehicle operation records and temperature and humidity sampling records to the corresponding transportation task.

[0060] When different data sources do not have a unified transportation task identifier, read the vehicle identifier, transportation origin, transportation destination, and transportation time interval; group data records with the same vehicle identifier, the same transportation origin and transportation destination, and whose collection time is within the same transportation time interval into data records corresponding to the same transportation task.

[0061] When the statistical period for economic, social, and environmental data or industry development basic data is longer than the target period, the most recently published statistical data that includes the target period is read, and the corresponding statistical version is recorded. Vehicle operation data, temperature and humidity sensor data, and transportation OD data are aggregated into the target period according to the data generation time.

[0062] Through the above processing, a multi-source heterogeneous cold chain dataset for the target area within the target period is generated. The multi-source heterogeneous cold chain dataset includes the data source, statistical version, or collection time of each data record, and serves as the data input for step S102.

[0063] S102: Preprocess the multi-source heterogeneous cold chain data. The preprocessing includes cleaning, index calculation and dimensionless processing.

[0064] Cleaning, dimensionless conversion, and index calculation can be performed in different orders depending on the data type. For statistical data that can directly generate index values, dimensionless conversion is performed after cleaning. For indices that need to be calculated from original vehicle records, sensor records, or transportation OD records, dimensionless conversion is performed on the calculated index values ​​after cleaning and index calculation.

[0065] Data cleaning includes field standardization, unit unification, duplicate record handling, missing record handling, and handling of records that do not meet preset data validity conditions.

[0066] Field standardization is used to convert fields with the same meaning from different data sources into a unified field name. For example, license plate numbers, vehicle codes, and platform vehicle numbers can be uniformly mapped to vehicle identifiers.

[0067] Unit standardization is used to convert similar data into the same unit of measurement. For example, mileage is standardized to kilometers, running time is standardized to hours, and cold storage capacity is standardized to ten thousand tons.

[0068] For vehicle operation records or transportation OD records that are missing vehicle identification, collection time, transportation origin or transportation destination, the corresponding records will be marked as invalid records and will not be used for subsequent indicator calculations.

[0069] For duplicate records with the same vehicle identifier, the same collection time, and the same data type, the records are selected according to the pre-established data source priority rules.

[0070] The data source priority rules include data source identifier, acquisition device type, calibration status, number of complete fields, data update time, and priority order. First, it compares whether the acquisition device has valid calibration records. If the calibration status is the same, it compares whether the record simultaneously contains vehicle identifier, acquisition time, and corresponding indicator fields. If the number of complete fields is the same, it compares whether the data update time is within the target period. The record corresponding to the data source ranked first is used as the cleaned data record.

[0071] 1 Data acquisition device calibration status With valid calibration records 2 Key field integrity It also includes vehicle identification, data collection time, and indicator fields. 3 Data update time The update time is within the target period. 4 Data source version Release history with newer version numbers

[0072] Table 1 Data Source Priority Rules

[0073] The preset data validity conditions for temperature and humidity sensor data are determined based on the sensor range, sampling period, and platform field rules. Records will be marked as invalid if the temperature value exceeds the corresponding sensor range, the acquisition time cannot be parsed, the vehicle identifier is missing, or the duration of consecutive repetition of the same sample value exceeds the upper limit of the sensor's normal reporting cycle.

[0074] The preset data validity conditions for vehicle operation data are determined based on the vehicle positioning terminal's range and the platform's operating rules. When the mileage change per unit time exceeds the maximum permissible driving distance for the vehicle under the corresponding road conditions, the running time is negative, or the timestamps of adjacent records are reversed, the corresponding record will be marked as invalid.

[0075] See Figure 2 The evaluation index system includes a criterion layer, an element layer, and an indicator layer. The evaluation index system is shown in Table 2.

[0076] Economic and social environment A Economic level A1 GDP per capita A11 Economic and social environment A Population size A2 resident population A21 Economic and social environment A Industrial Structure A3 The proportion of the tertiary industry A31 Economic and social environment A Industrial Structure A3 Total output value of agriculture, forestry, animal husbandry and fishery A32 Economic and social environment A Consumption level A4 per capita disposable income A41 Economic and social environment A Consumption level A4 Total retail sales of consumer goods A42 Economic and social environment A Information technology level A5 Mobile phone users A51 Economic and social environment A Information technology level A5 Internet broadband access numbers A52 Industry development level B Traffic conditions B1 Highway mileage B11 Industry development level B Traffic conditions B1 Freight volume B12 Industry development level B Facility Supply B2 Cold storage capacity B21 Industry development level B Facility Supply B2 Number of refrigerated trucks B22 Industry development level B Market Entity B3 Number of cold chain logistics companies B31 Industry development level B Market Entity B3 Number of transportation workers B32 Industry development level B Greening level B4 The proportion of new energy refrigerated trucks B41 Industry operating status C Operational scale C1 Vehicle operating rate C11 Industry operating status C Operational scale C1 Number of transport vehicles C12 Industry operating status C Operating efficiency C2 Average daily mileage of vehicles C21 Industry operating status C Operating efficiency C2 Average daily running time of vehicles C22 Industry operating status C Temperature control safety C3 Temperature control compliance rate throughout the process C31 Industry operating status C Temperature control safety C3 Vehicle over-temperature alarm rate C32 Industry operating status C Temperature control safety C3 Number of vehicle over-temperature alarms C33 Network Development Level D Spatial structure D1 Spatial concentration D11 Network Development Level D Network topology D2 Network density D21 Network Development Level D Network topology D2 Network clustering D22 Network Development Level D Network topology D2 Network connectivity efficiency D23 Network Development Level D Node Function D3 Node connectivity D31 Network Development Level D Node Function D3 Network Intermediation D32 Network Development Level D Node Function D3 Connectivity D33

[0077] Table 2 Evaluation Index System for Cold Chain Transportation

[0078] Vehicle operating rate is calculated using the following formula:

[0079]

[0080] in, The number of vehicles engaged in cold chain transportation activities within the target period; This refers to the number of refrigerated trucks connected to the road cold chain transportation platform within the target period.

[0081] The number of transport trips refers to the number of cold chain transport trips completed within the target period. A transport task is counted as one transport trip from the start of the transport to the destination.

[0082] The average daily mileage of a vehicle is calculated using the following formula:

[0083]

[0084] in, For the first The mileage corresponding to each effective vehicle day. This refers to the number of valid vehicles per day within the target period.

[0085] The average daily running time of a vehicle is calculated using the following formula:

[0086]

[0087] in, For the first The running time corresponding to each valid vehicle day.

[0088] The overall temperature control compliance rate is calculated using the following formula:

[0089]

[0090] in, The cumulative running time during which the temperature of the corresponding goods remains within the preset temperature control range within the target period; This refers to the cumulative runtime of the target period with valid temperature sampling records.

[0091] The preset temperature control range is determined based on the industry standards corresponding to the transported goods category, the temperature control range recorded in the transport contract, or the configuration parameters of the corresponding goods category in the road cold chain transport platform. When the current transport task is executed, the corresponding preset temperature control range is retrieved according to the goods category in the transport task record, and each temperature sampling value is compared with the lower limit and upper limit of the preset temperature control range, respectively.

[0092] The vehicle over-temperature alarm rate is calculated using the following formula:

[0093]

[0094] in, The number of vehicles that trigger at least one over-temperature alarm within the target period.

[0095] The number of times a vehicle over-temperature alarm is triggered is calculated using the following formula:

[0096]

[0097] in, This represents the total number of over-temperature alarm events within the target period.

[0098] When the time interval between adjacent over-temperature sampling records is no greater than the preset event merging time, the adjacent over-temperature sampling records are merged into the same over-temperature alarm event; when the time interval is greater than the preset event merging time, they are counted as different over-temperature alarm events. The preset event merging time is set as the sum of the temperature and humidity sensor reporting cycle and the allowed continuous packet loss cycle of the road cold chain transportation platform.

[0099] Let the number of evaluation samples be... The number of evaluation indicators is , No. The evaluation sample at the ... The original indicator value for each indicator is .

[0100] For positive indicators where a larger value signifies better overall operating conditions for cold chain transportation, the following formula is used for dimensionless transformation:

[0101]

[0102] For the inverse indicator that a smaller value indicates better overall operating conditions for cold chain transportation, the following formula is used for forward consistency and dimensionlessness:

[0103]

[0104] in, and The first The maximum and minimum values ​​of each indicator in all evaluation samples; The index value is the dimensionless value, and It is located between 0 and 1.

[0105] when When, it indicates the first There are no differences among the indicators in the current evaluation sample. If there is a confirmed industry benchmark value, the corresponding dimensionless indicator value is determined based on the industry benchmark value; if there is no industry benchmark value, the information content of the indicator is set to zero so that the indicator does not increase the CRITIC objective weight.

[0106] Step S102 yields dimensionless index values ​​corresponding to economic and social environmental indicators, industry development indicators, and industry operation indicators, and retains the cleaned transportation OD data for step S103 to calculate network development indicators.

[0107] S103: Construct a complex weighted network based on transportation OD data and calculate network development indicators.

[0108] See Figure 3 The network is constructed by taking the origin and destination of transportation or the node areas divided according to administrative divisions and cold chain logistics nodes as network nodes, taking the cold chain transportation links between different network nodes as network edges, and taking the transportation volume, transportation frequency or number of transportation vehicles as the network edge weights.

[0109] When it is necessary to retain the transportation direction, a directed weighted network is constructed; when only the existence of transportation links between nodes is evaluated without distinguishing the transportation direction, an undirected weighted network is constructed. Different evaluation samples in the same evaluation batch use the same network node partitioning method, network direction type, and network edge weight type.

[0110] Let the number of network nodes in the complex weighted network be . , No. The transportation volume or frequency of each network node is The total network traffic volume or total traffic frequency is .

[0111] Spatial concentration is calculated according to the following formula:

[0112]

[0113] in:

[0114]

[0115] When there are no transportation records within the target period, At that time, the spatial concentration of the corresponding evaluation sample is not calculated, and the evaluation sample is marked as lacking transport OD data.

[0116] For undirected networks, the network density is calculated using the following formula:

[0117]

[0118] For directed networks, the network density is calculated using the following formula:

[0119]

[0120] in, This represents the actual number of network edges in a complex weighted network.

[0121] Network clustering is calculated using the following formula:

[0122]

[0123] For an undirected network, nodes The clustering coefficient is:

[0124]

[0125] in, For nodes The actual number of network edges between adjacent nodes; For nodes The number of adjacent nodes. When At that time, Set to zero.

[0126] For directed weighted networks, the network clustering degree is calculated after converting them into adjacency projection networks that do not distinguish directions. The same conversion method is used for the same evaluation batch.

[0127] Network connectivity efficiency is characterized by the average shortest path length. For a pair of network nodes with reachable paths, the average shortest path length is calculated using the following formula:

[0128]

[0129] in, For network nodes To network node The shortest path length between them; This represents the number of network node pairs with reachable paths.

[0130] A smaller average shortest path length indicates higher network connectivity efficiency. Therefore, when forming the evaluation index set, the average shortest path length is used as a reverse index for positive consistency.

[0131] When a complex weighted network contains multiple connected components, the average shortest path length of each connected component is calculated separately, and then the number of reachable network nodes in each connected component is weighted and summed.

[0132] Node connectivity is represented by the sum of the weights of the network edges connected to the corresponding network node:

[0133]

[0134] in, For nodes The set of adjacent nodes; For nodes and nodes The weights of the network edges between them.

[0135] The intermediary degree of a node network is calculated using the following formula:

[0136]

[0137] in, For nodes To the node The number of shortest paths between them; For nodes To the node Between nodes The number of shortest paths. When node With nodes When there is no path between them, the corresponding node pairs are not included in the summation.

[0138] For a connected network, the ease of node connectivity is calculated using the following formula:

[0139]

[0140] The greater the connectivity of a node, the shorter the path length required for the corresponding network node to reach other network nodes.

[0141] For disconnected networks, the connectivity advantages of near-central computing nodes are utilized:

[0142]

[0143] For unreachable network node pairs, set the corresponding inverse to zero.

[0144] The spatial concentration, network density, network clustering, network connectivity efficiency, node connectivity, network intermediation, and connectivity convenience are made dimensionless and then combined with the economic and social environmental indicators, industry development indicators, and industry operation indicators obtained in step S102 to generate a complete set of evaluation indicators. This set of evaluation indicators is used in step S104 to determine the combined weights.

[0145] S104: Determine the combined weights of each indicator in the evaluation indicator set.

[0146] See Figure 4 Combinatorial weighting includes subjective weighting based on the analytic hierarchy process (AHP), objective weighting based on the CRITIC method, and combined weighting based on game theory.

[0147] 1. Subjective weighting in the Analytic Hierarchy Process (AHP).

[0148] Judgment matrices are constructed for the criterion layer, element layer, and indicator layer, respectively. Evaluation objects within the same judgment matrix are compared pairwise by experts or evaluators involved in the cold chain transportation evaluation, according to a pre-defined importance scale.

[0149] A 1-9 scale is used. A value of 1 indicates that the two evaluated objects have the same importance. Values ​​of 3, 5, 7, and 9 indicate that the former evaluated object has slightly stronger, stronger, stronger, and extremely stronger importance relative to the latter evaluated object, respectively. Values ​​of 2, 4, 6, and 8 indicate the intermediate state between adjacent judgment levels, and the reciprocal indicates the result of the reverse comparison.

[0150] Let the judgment matrix be... for:

[0151]

[0152] in, To determine the order of a matrix; For the first The evaluation object is relative to the first The importance scale value of each evaluation object For the first Each evaluated object has an importance scale value relative to itself, and satisfies the following:

[0153]

[0154] Find the largest eigenvalue of the judgment matrix The corresponding feature vectors are then normalized to obtain local subjective weights.

[0155] Consistency Indicators Calculate according to the following formula:

[0156]

[0157] Consistency ratio Calculate according to the following formula:

[0158]

[0159] in, This is a random consistency index corresponding to the order of the judgment matrix. When... If the condition is met, the judgment matrix is ​​confirmed to have passed the consistency test; otherwise, the importance scale values ​​in the judgment matrix are readjusted and the consistency test is performed again.

[0160] Set the criteria layer The weight relative to the overall goal is Element layer Relative to the criterion layer The weight is ,index Relative to the feature layer The weight is Then the indicator Global subjective weight for:

[0161]

[0162] The subjective weight vector is composed of the global subjective weights of all indicators. :

[0163]

[0164] 2. CRITIC method of objective weighting.

[0165] The CRITIC method calculates objective weights based on evaluation samples from multiple regions or multiple periods. Let the number of evaluation samples be... The number of evaluation indicators is The dimensionless index value is .

[0166] No. Sample mean of each indicator for:

[0167]

[0168] No. Standard deviation of each indicator for:

[0169]

[0170] No. The first indicator and the first Correlation coefficients among indicators for:

[0171]

[0172] No. Information content of each indicator for:

[0173]

[0174] No. Objective weight of each indicator for:

[0175]

[0176] The objective weight vector is composed of all objective weights. :

[0177]

[0178] When the value of a certain indicator is the same across all evaluation samples, the information content of that indicator is set to zero. When the sum of the information content of all indicators is zero, the subjective weight vector is used as the subsequent combined weight vector.

[0179] 3. Game theory combinatorial weighting.

[0180] Subjective weight vector and objective weight vector As the basic weight vector.

[0181] Game theory-based combinatorial weighting aims to achieve equilibrium by resolving the sum of squared deviations between the combined weight vector and each of the basic weight vectors. This sum of squared deviations characterizes the overall deviation of the combined weight vector from both subjective and objective weight vectors. The initial combined coefficients for each basic weight vector are obtained by solving the combinatorial coefficient equation corresponding to the optimization objective.

[0182] Combined weight vector Represented as:

[0183]

[0184] The equation for the combination coefficients is:

[0185]

[0186] Written in matrix form:

[0187]

[0188] The obtained initial combination coefficients are then nonnegatively normalized:

[0189]

[0190] Final combined weight vector for:

[0191]

[0192] In the formula, It is the index of the basic weight vector, and ∈{1,2}; Let be the index of the equation with combined coefficients, and ∈{1,2} Indicates the first A basic weight vector, Indicates the first A basic weight vector; superscript Indicates vector transpose; Indicates the first The fundamental weight vector and the first The inner product of the basic weight vectors; Represents the initial combination coefficients The absolute value of.

[0193] When the coefficient matrix of the combined coefficient equation is a singular matrix, the generalized inverse is used to solve it; when the subjective weight vector and the objective weight vector are the same, the subjective weight vector is directly determined as the combined weight vector.

[0194] In step S104, combined weights of each indicator in the evaluation indicator set are generated. These combined weights are used in step S105 to calculate the comprehensive evaluation cloud and in step S106 to calculate the obstacle degree of the indicator layer.

[0195] S105: Determining the evaluation level of cold chain transportation operations based on cloud models

[0196] See Figure 5 The cloud model evaluation includes the construction of a pre-set grade standard cloud, the determination of cloud digital features for each indicator, and the comprehensive evaluation of cloud computing and cloud similarity matching.

[0197] The preset cloud-based level standards include:

[0198] The operational status of cold chain transportation is divided into five levels: low level, relatively low level, medium level, relatively high level, and high level.

[0199] After the indicators are positively consistent and dimensionless, they fall within the range. At that time, the fixed interval method was used to determine the evaluation level intervals shown in Table 3.

[0200] low level lower level medium level High level High level

[0201] Table 3 Evaluation Level Range

[0202] In other implementations, five evaluation level boundaries are determined based on the quantiles of multiple historical evaluation samples or established industry evaluation benchmarks. All evaluation samples in the same evaluation batch adopt the same evaluation level boundaries.

[0203] Let the first The rating level range is as follows: The expected level of the cloud corresponds to the preset standard level. Entropy Calculate according to the following formulas respectively:

[0204]

[0205]

[0206] hyperentropy Determined based on historical evaluation samples corresponding to the evaluation level. Let the [number]th [level] be... The evaluation levels include There are 1 historical sample, and the historical sample value is 1. ,but:

[0207]

[0208]

[0209]

[0210] in, For the first The sample variance of historical samples for each rating level.

[0211] When historical samples for the corresponding evaluation level are lacking, the hyperentropy is determined according to the following formula:

[0212]

[0213] in, A positive coefficient less than 1. The aforementioned... The determination is based on the dispersion of samples of the same grade in existing historical evaluation results, and remains unchanged within the same evaluation batch.

[0214] A comprehensive evaluation of cloud computing includes:

[0215] Let the first The cloud digital characteristics corresponding to each indicator are: The corresponding combination weight is ,and:

[0216]

[0217] Overall evaluation of cloud expectations for:

[0218]

[0219] Comprehensive evaluation of cloud entropy for:

[0220]

[0221] Comprehensive evaluation of cloud hyperentropy for:

[0222]

[0223] This leads to a comprehensive evaluation of cloud. :

[0224]

[0225] Cloud similarity matching includes:

[0226] The corresponding cloud digital characteristics of the comprehensive evaluation cloud and the five preset level standard clouds are standardized using Min-Max.

[0227] Set up a standardized comprehensive evaluation cloud for:

[0228]

[0229] No. A preset level standard cloud for:

[0230]

[0231] Comprehensive evaluation of cloud and the first The Euclidean distance between the preset standard cloud levels is:

[0232]

[0233] Corresponding cloud similarity for:

[0234]

[0235] By comparing the cloud similarity of the comprehensive evaluation cloud with the cloud corresponding to the five preset level standard clouds, the evaluation level corresponding to the cloud with the highest similarity is determined as the evaluation level of cold chain transportation operation.

[0236] When at least two rating levels correspond to the same cloud similarity, the expected value of reading the comprehensive rating cloud is... The evaluation level corresponding to the evaluation interval where the expectation is located is determined as the cold chain transportation operation evaluation level; when the expectation is located at the common boundary of two evaluation intervals, the two adjacent evaluation levels and their cloud similarity are written into the processing result at the same time.

[0237] In step S105, a cold chain transportation operation evaluation level and a comprehensive evaluation cloud are generated, along with the cloud similarity to the preset level standard clouds. The cold chain transportation operation evaluation level is then used in step S106, along with the obstacle factor ranking results, to form the final processing result.

[0238] S106: Calculate the obstacle degree and generate the obstacle factor ranking results.

[0239] Let the target evaluation sample be the first The index values ​​after positive uniformity and dimensionless transformation are: , No. The combined weight of each indicator is .

[0240] No. Contribution of each indicator for:

[0241]

[0242] No. The deviation of each indicator for:

[0243]

[0244] No. Indicator layer barrier for each indicator for:

[0245]

[0246] when:

[0247]

[0248] When all evaluation indicators have reached the normalization upper limit in the current evaluation sample, the barrier degree of all indicator layers is set to zero, and the ranking results of barrier-free factors are output.

[0249] Set up element layer Includes a set of indicators Corresponding element layer obstacle degree for:

[0250]

[0251] Set the criteria layer Includes feature layer set Corresponding criterion layer obstacle degree for:

[0252]

[0253] The obstacle factors are sorted in descending order of obstacle degree at the indicator layer, element layer, and criterion layer, respectively, to generate the obstacle factor ranking results.

[0254] The ranking results of the obstacle factors include the obstacle factor name, the criterion layer to which it belongs, the element layer to which it belongs, the degree of obstacle, and the ranking position.

[0255] The cold chain transportation operation evaluation level generated in step S105 is read, and the cold chain transportation operation evaluation level and the obstacle factor ranking result are combined to form the cold chain transportation multi-source operation data processing result. The processing result also includes the dimensionless index value of each indicator, the combined weight, the comprehensive evaluation cloud digital characteristics, and the cloud similarity corresponding to each preset level standard cloud.

[0256] The basis for setting each preset parameter and the calling method are shown in Table 4.

[0257] Target period Evaluation task cycle and data update cycle Used for aggregating multi-source heterogeneous cold chain data Data validity conditions Sensor range, field rules and sampling period Used to mark invalid records Temperature control range Goods category standards, transportation contracts or platform configuration Used to calculate the duration of temperature control compliance. Event merging duration Sensor reporting cycle and allowable continuous packet loss cycle Used for merging over-temperature sampling records AHP Importance Scale 1 to 9 scale method Used to construct the judgment matrix Consistency threshold AHP Consistency Verification Rules Example value is 0.1 Evaluation level boundaries Fixed interval, historical sample quantiles or industry benchmarks Used to build a cloud with a preset level of standard. hyperentropy coefficient Dispersion of historical samples at the same level Used to determine hyperentropy when there are no historical samples Network node partitioning Administrative regions, logistics nodes, or transportation origin and destination points Used to construct complex weighted networks Network edge weight types Transport volume, transport frequency, or number of transport vehicles Used to determine network edge weights

[0258] Table 4 Preset Parameter Setting Method

[0259] Example 2

[0260] The following numerical examples are used to illustrate the processing procedure of this application and do not constitute a limitation on the number of evaluation indicators, indicator values, combined weights, or evaluation results.

[0261] Five evaluation samples were selected from the target region R0 and the comparison regions R1 to R4, corresponding to the data in the same year. For ease of explanation, Table 5 lists eight indicators: GDP per capita, cold storage capacity, vehicle operating rate, average daily vehicle mileage, overall temperature control compliance rate, vehicle over-temperature alarm rate, network density, and network connectivity efficiency. The vehicle over-temperature alarm rate and the average shortest path length, used to characterize network connectivity efficiency, have been forward-aligned according to the inverse indicators.

[0262] R0 0.72 0.65 0.68 0.60 0.75 0.55 0.62 0.58 R1 0.55 0.50 0.60 0.48 0.62 0.40 0.50 0.45 R2 0.80 0.75 0.78 0.70 0.82 0.68 0.72 0.66 R3 0.45 0.58 0.52 0.55 0.57 0.35 0.46 0.50 R4 0.68 0.62 0.66 0.64 0.70 0.50 0.60 0.56

[0263] Table 5. Dimensionless results for some indicators

[0264] The subjective weights of the eight indicators were obtained using the analytic hierarchy process (AHP). :

[0265]

[0266] The standard deviation and correlation coefficient were calculated based on the five evaluation samples shown in Table 5 to obtain the objective weights. :

[0267]

[0268] The normalized combinatorial coefficients are obtained through game theory combinatorial weighting:

[0269]

[0270]

[0271] Obtain the combined weights :

[0272]

[0273] Based on the index value of the target region R0 and the combined weights, the comprehensive evaluation cloud expectation is obtained. :

[0274]

[0275] Comprehensive evaluation cloud for:

[0276]

[0277] The five preset standard cloud levels are as follows:

[0278]

[0279]

[0280]

[0281]

[0282]

[0283] The calculated cloud similarities between the comprehensive evaluation cloud and the five preset standard clouds were 0.6455, 0.7411, 0.8701, 0.9513, and 0.7994, respectively. The highest cloud similarity of 0.9513 corresponds to a higher-level standard cloud, therefore the cold chain transportation operation evaluation level of the target area R0 was determined to be higher.

[0284] Further calculations were performed on the indicator layer barrier degree of the eight indicators, resulting in the results shown in Table 6.

[0285] 1 Average daily mileage of vehicles C21 17.43% 2 Cold storage capacity B21 14.25% 3 Vehicle over-temperature alarm rate C32 13.18% 4 Network connectivity efficiency D23 12.23% 5 GDP per capita A11 12.10% 6 Vehicle operating rate C11 11.86% 7 Network density D21 10.01% 8 Full-process temperature control compliance rate C31 8.93%

[0286] Table 6. Obstacles to Some Indicators

[0287] The obstacle degrees of each indicator layer are summarized according to the element layer and criterion layer in Table 2 to obtain the obstacle degrees of the element layer such as operating efficiency, facility supply and temperature control safety, as well as the obstacle degrees of the criterion layer corresponding to the economic and social environment, industry development level, industry operation status and network development level.

[0288] Example 3

[0289] See Figure 6 This embodiment provides a multi-source operation data processing system for cold chain transportation, including a data acquisition module, an indicator processing module, a network construction module, a combined weighting module, a cloud model evaluation module, an obstacle identification module, and a result output module.

[0290] The data acquisition module is used to acquire economic and social environmental data, industry development basic data, vehicle operation data, temperature and humidity sensor data, and transportation OD data of the target area within the target period, and to establish the correlation between different data records based on vehicle identification, transportation task identification, and collection time.

[0291] The indicator processing module is used to standardize fields, unify units, process duplicate records, process missing records, and process invalid records for multi-source heterogeneous cold chain data. It calculates economic and social environmental indicators, industry development indicators, and industry operation indicators, and performs positive consistency and dimensionless processing on each indicator.

[0292] The network construction module is used to determine network nodes, network edges, and network edge weights based on transportation OD data, construct directed weighted networks or undirected weighted networks, and calculate spatial concentration, network density, network clustering, network connectivity efficiency, node connectivity, network intermediation, and connectivity convenience.

[0293] The combined weighting module is used to calculate subjective weights using the analytic hierarchy process (AHP), objective weights using the CRITIC method, and combined weights using game theory combined weighting.

[0294] The cloud model evaluation module is used to determine the preset level standard cloud based on the evaluation level range or historical evaluation samples, generate a comprehensive evaluation cloud based on the combined weights of each indicator and cloud digital characteristics, and determine the cold chain transportation operation evaluation level based on the similarity between the comprehensive evaluation cloud and each preset level standard cloud.

[0295] The obstacle identification module is used to calculate the indicator deviation based on the dimensionless indicator value, calculate the indicator layer obstacle degree based on the indicator deviation and combined weight, and obtain the element layer obstacle degree and criterion layer obstacle degree by summarizing the hierarchical affiliation in the evaluation indicator system.

[0296] The results output module is used to write the cold chain transportation operation evaluation level, obstacle factor ranking results, dimensionless index values, combined weights, comprehensive evaluation cloud digital features, and cloud similarity corresponding to each preset level standard cloud into the cold chain transportation multi-source operation data processing results, and output them through display terminals, data interfaces, or report files.

[0297] Each module in this embodiment can be implemented by a software program, or by a processor, memory, communication interface, and data storage device working together. When the processor executes the program stored in the memory, it implements the aforementioned multi-source operation data processing method for cold chain transportation.

[0298] The above description is merely a preferred embodiment of this application. It should be noted that, due to the limitations of written expression, there are objectively infinite specific structures. For those skilled in the art, without departing from the principles of this application, several improvements, modifications, or variations can be made, and the above technical features can be combined in an appropriate manner. These improvements, modifications, variations, or combinations, or the direct application of the concept and technical solution of this application to other situations without modification, should all be considered within the scope of protection of this application.

Claims

1. A method for processing multi-source operational data in cold chain transportation, characterized in that, include: Acquire multi-source heterogeneous cold chain data for the target area within the target period; The multi-source heterogeneous cold chain data is cleaned, dimensionlessized, and indexed to obtain an evaluation index set that includes economic, social, and environmental indicators, industry development indicators, industry operation indicators, and network development indicators. The combined weights of each indicator in the evaluation index set are determined by a combination of the analytic hierarchy process, the CRITIC method, and game theory. A comprehensive evaluation cloud is generated based on the combined weights and cloud digital features of each indicator, and the comprehensive evaluation cloud is matched with the preset level standard cloud to obtain the cold chain transportation operation evaluation level. The deviation of an indicator is determined based on the dimensionless values ​​of each indicator in the evaluation indicator set. The obstacle degree of the indicator layer, the element layer, and the criterion layer is calculated based on the deviation of the indicator and the combined weight. The obstacle factor ranking results are generated according to the obstacle degree. The output includes the cold chain transportation operation evaluation level and the ranking results of the obstacle factors, which are the multi-source operation data processing results of cold chain transportation.

2. The method for processing multi-source operation data in cold chain transportation according to claim 1, characterized in that, The multi-source heterogeneous cold chain data includes economic, social and environmental data, industry development basic data, vehicle operation data, temperature and humidity sensor data, and transportation OD data; the economic, social and environmental data and the industry development basic data are obtained from macroeconomic statistical data and industry statistical data, while the vehicle operation data, the temperature and humidity sensor data, and the transportation OD data are obtained from the road cold chain transportation platform.

3. The method for processing multi-source operation data in cold chain transportation according to claim 2, characterized in that, Acquiring the multi-source heterogeneous cold chain data includes: reading the corresponding statistical versions of the economic, social and environmental data and the industry development basic data according to the target period; collecting the vehicle operation data, the temperature and humidity sensor data and the transportation OD data according to the data generation time; and establishing the correlation between different data records based on vehicle identification, transportation task identification and collection time.

4. The method for processing multi-source operation data in cold chain transportation according to claim 3, characterized in that, The evaluation index set is constructed according to the criteria layer, the element layer, and the index layer; the industry operation index includes vehicle operation rate, number of transport vehicles, average daily vehicle mileage, average daily vehicle running time, full-process temperature control compliance rate, vehicle over-temperature alarm rate, and number of vehicle over-temperature alarms. The network development indicators are obtained based on a complex weighted network constructed from the transportation OD data; the network development indicators include spatial concentration, network density, network clustering, network connectivity efficiency, node connectivity, network intermediation, and connectivity convenience.

5. The method for processing multi-source operation data in cold chain transportation according to claim 4, characterized in that, Constructing the complex weighted network based on the transport OD data includes: The starting point, ending point, or node area of ​​transportation are defined as network nodes, the transportation connections between different network nodes are defined as network edges, and the weight of the network edges is determined according to the transportation volume, transportation frequency, or number of transportation vehicles. A directed weighted network or an undirected weighted network is constructed based on whether the transportation direction is retained, and the network development index is determined based on the complex weighted network.

6. The method for processing multi-source operation data in cold chain transportation according to claim 1, characterized in that, The subjective weights are determined using the analytic hierarchy process (AHP), including: The criterion-level judgment matrix, the element-level judgment matrix, and the indicator-level judgment matrix are constructed respectively, and the evaluation objects at the same level are compared pairwise according to the preset importance scale. Consistency checks are performed on each judgment matrix, and when the consistency ratio is less than a preset consistency threshold, the local subjective weights of each layer are synthesized step by step to obtain the global subjective weights of each indicator in the evaluation index set.

7. The method for processing multi-source operation data in cold chain transportation according to claim 6, characterized in that, The CRITIC method is used to determine objective weights, including: Based on multiple regional evaluation samples or multiple periodic evaluation samples, including the evaluation samples corresponding to the target region and the target period, calculate the standard deviation of each indicator and the correlation coefficient between any two indicators. The information content of each indicator is determined based on the standard deviation and the correlation coefficient, and the information content of each indicator is normalized to obtain the objective weight of each indicator in the evaluation indicator set.

8. The method for processing multi-source operation data in cold chain transportation according to claim 7, characterized in that, The combination weights are determined using game theory-based combinatorial weighting, including: Using the global subjective weight and the objective weight as the basic weight vector, and taking the balance of the sum of squared deviations between the combined weight vector and each basic weight vector as the optimization objective, the combination coefficients corresponding to each basic weight vector are solved. The combination coefficients are normalized, and the basic weight vectors are linearly combined based on the normalized combination coefficients to obtain the combination weights.

9. The method for processing multi-source operation data in cold chain transportation according to claim 8, characterized in that, Determining the cold chain transportation operation evaluation level includes: The evaluation domain corresponding to the index values ​​after positive uniformity and dimensionlessness is divided into five evaluation levels: low level, lower level, medium level, higher level and high level. The expected value and entropy of the corresponding preset level standard cloud are determined based on the upper and lower boundaries of each evaluation level interval, and the super-entropy is determined based on the historical evaluation samples of the corresponding evaluation level. When the historical evaluation samples are missing, the super-entropy is determined based on the preset super-entropy coefficient and the entropy. The cloud digital features of the comprehensive evaluation cloud and each preset level standard cloud are standardized, the Euclidean distance between the comprehensive evaluation cloud and each preset level standard cloud is calculated, the cloud similarity is determined based on the Euclidean distance, and the evaluation level corresponding to the maximum cloud similarity is determined as the cold chain transportation operation evaluation level.

10. A multi-source operation data processing system for cold chain transportation, implementing the multi-source operation data processing method for cold chain transportation as described in any one of claims 1 to 9, characterized in that, include: The data acquisition module is used to acquire multi-source heterogeneous cold chain data of the target area within the target period; The indicator processing module is used to clean, dimensionlessize, and calculate indicators for the multi-source heterogeneous cold chain data to obtain an evaluation indicator set that includes economic and social environmental indicators, industry development indicators, industry operation indicators, and network development indicators. The combined weighting module is used to determine the combined weights of each indicator in the evaluation indicator set by combining the analytic hierarchy process, the CRITIC method and game theory. The cloud model evaluation module is used to generate a comprehensive evaluation cloud based on the combined weights and cloud digital features of each indicator, and to perform similarity matching between the comprehensive evaluation cloud and the preset level standard cloud to obtain the cold chain transportation operation evaluation level. The obstacle identification module is used to determine the deviation degree of each indicator in the evaluation index set after dimensionless transformation, calculate the obstacle degree of the indicator layer, the element layer and the criterion layer according to the deviation degree and the combined weight, and generate the obstacle factor ranking result according to the obstacle degree. The results output module is used to output the multi-source operation data processing results of cold chain transportation, including the cold chain transportation operation evaluation level and the obstacle factor ranking results.

Images