Cold chain transportation demand analysis method and system based on big data prediction

Abstract

The invention provides a cold-chain transportation demand analysis method and system based on big data prediction, and the method comprises the steps: firstly obtaining a historical cold-chain order data set which comprises temperature-sensitive cargo transportation records of multiple time periods and corresponding environment parameters in a target region, and then extracting a transportation demand feature set from the historical cold-chain order data set; the characteristics of correlation between cargo types and temperature sensitivity, balance between transportation aging and equipment energy consumption, matching between regional climate and storage conditions and the like are covered. And then inputting the feature set into a pre-trained space-time prediction model, and generating multi-dimensional cold chain demand parameters, such as a temperature fluctuation tolerance threshold value and the like, of the target area in a specified period in the future. And finally, based on the multi-dimensional cold chain demand parameters, generating a cold chain transportation dynamic scheduling scheme with a timestamp. According to the method, the cold chain transportation requirements can be accurately analyzed by integrating multi-dimensional factors, scientific dynamic scheduling is realized, and the cold chain transportation efficiency and quality are improved.

Description

Technical Field

[0001] The present application relates to the field of cold chain transportation technology, and specifically, to a cold chain transportation demand analysis method and system based on big data prediction. Background Art

[0002] In the development history of the cold chain transportation industry, the traditional cold chain transportation demand analysis method faces many limitations and is unable to meet the growing market demand and refined management requirements.

[0003] In the early days, cold chain transportation mainly relied on manual experience and simple statistical data to estimate transportation demand. Operators roughly judged the transportation task volume in different periods based on past work experience and limited historical order data. However, this method lacked systematicity and scientificity, and could not accurately respond to the complex and changing market environment and diversified cargo demand. Due to the lack of comprehensive data support, the special transportation requirements of different goods in different environments were often underestimated, resulting in problems such as improper temperature control during transportation, affecting the quality of the goods and causing economic losses.

[0004] With the development of technology, some transportation platforms have begun to use relatively simple data statistics methods to analyze transportation demand. These methods usually only focus on data of a single dimension, such as simply counting the number of orders to predict future transportation volume, while ignoring many factors closely related to cold chain transportation. For example, the complex internal relationship between the type of goods and temperature sensitivity is not taken into account. Different temperature-sensitive goods have very different temperature requirements during transportation. Simple data statistics cannot accurately grasp the appropriate transportation temperature range for each type of goods, making it difficult to ensure the quality of the goods during transportation.

[0005] In terms of transportation timeliness and equipment energy consumption, traditional analysis methods have failed to fully recognize the balance between the two. Often, in pursuit of transportation timeliness, equipment energy is over-consumed, resulting in a significant increase in operating costs; or in order to save energy, transportation timeliness is sacrificed, affecting the timeliness of goods delivery and customer satisfaction. At the same time, previous methods mostly considered regional climate and storage conditions in isolation, without recognizing the important impact of regional climate conditions on the setting of storage conditions, and the key role of the matching degree between the two in the effect of cold chain transportation. This makes storage construction and operation lack of pertinence, unable to effectively adapt to the climate characteristics of different regions, and further increases the difficulty and cost of cold chain transportation.

[0006] Existing forecasting models also have obvious deficiencies. Most of them are based on simple time series analysis or conventional machine learning models, and do not fully consider the complex changes in cold chain transportation demand in time and space. These models cannot accurately capture the dynamic differences in cold chain demand between different time periods and regions, resulting in a large deviation between the forecast results and actual demand, making it difficult to guide effective transportation scheduling.

[0007] Traditional cold chain transportation scheduling plans are usually static and rarely adjusted according to actual conditions once formulated. This static scheduling method cannot cope with sudden changes in demand during transportation, nor can it be optimized according to real-time transportation conditions and environmental factors, resulting in unreasonable resource allocation and low transportation efficiency. Summary of the invention

[0008] In view of this, the purpose of this application is to provide a cold chain transportation demand analysis method and system based on big data prediction.

[0009] According to the first aspect of the present application, a cold chain transportation demand analysis method based on big data prediction is provided, the method comprising: Acquire a historical cold chain order dataset of a target area, wherein the historical cold chain order dataset includes temperature-sensitive cargo transportation records and their corresponding environmental parameters over multiple time periods; Extracting a set of transport demand features from the historical cold chain order data, wherein the set of transport demand features includes a correlation feature between the type of goods and temperature sensitivity, a balance feature between transport timeliness and equipment energy consumption, and a matching feature between regional climate and storage conditions; Input the transport demand feature set into a pre-trained spatiotemporal prediction model to generate multi-dimensional cold chain demand parameters for the target area within a future specified period, wherein the multi-dimensional cold chain demand parameters include a temperature fluctuation tolerance threshold, a transport timeliness critical value, and a path node load coefficient; Based on the multi-dimensional cold chain demand parameters, a dynamic scheduling plan for cold chain transportation with a timestamp is generated.

[0010] According to the second aspect of the present application, a cold chain logistics service system is provided, which includes a machine-readable storage medium and a processor, wherein the machine-readable storage medium stores machine-executable instructions, and when the processor executes the machine-executable instructions, the cold chain logistics service system implements the aforementioned cold chain transportation demand analysis method based on big data prediction.

[0011] According to the third aspect of the present application, a computer-readable storage medium is provided, in which computer-executable instructions are stored. When the computer-executable instructions are executed, the aforementioned cold chain transportation demand analysis method based on big data prediction is implemented.

[0012] According to any one of the above aspects, the technical effect of the present application is: The embodiment of the present application obtains a historical cold chain order data set of temperature-sensitive cargo transportation records and their corresponding environmental parameters in the target area for multiple time periods, associates the transportation records with the environmental parameters, and defines a set of transportation demand feature sets, among which the association feature between cargo type and temperature sensitivity breaks the previous simple classification of cargo temperature requirements, deeply explores the intrinsic connection between different cargoes in temperature sensitivity, and provides a basis for more accurate control of transportation temperature; the balance feature between transportation timeliness and equipment energy consumption solves the long-standing problem of balancing timeliness and energy consumption in cold chain transportation, and provides a new perspective for optimizing transportation strategies by analyzing the relationship between the two; the matching feature between regional climate and storage conditions combines the external environment with the internal storage conditions, which is a key connection overlooked in previous cold chain transportation demand analysis and helps to rationally plan storage layout and condition settings.

[0013] In the prediction stage, the extracted transport demand feature set is input into the pre-trained spatiotemporal prediction model to generate multi-dimensional cold chain demand parameters for the target area within the specified future period. The generation of multi-dimensional parameters such as temperature fluctuation tolerance threshold, transport timeliness critical value and path node load coefficient can comprehensively and forward-lookingly characterize the cold chain transport demand compared to the traditional single-dimensional prediction, allowing transport companies to understand in advance the future transportation process in terms of temperature control, time scheduling, resource allocation and other requirements.

[0014] Traditional scheduling solutions often lack comprehensive consideration of time dimensions and multi-dimensional demand parameters, and are difficult to adapt to the complex and changing cold chain transportation environment. However, this solution, by combining timestamps and multi-dimensional parameters, can scientifically schedule transportation needs at different times in the future in real time and dynamically, thereby improving the utilization efficiency of cold chain transportation resources, reducing transportation costs, and maximizing the transportation quality of temperature-sensitive goods, thus improving the intelligence and adaptability of the entire cold chain transportation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 A schematic diagram of the process of the cold chain transportation demand analysis method based on big data prediction provided in an embodiment of the present application is shown; Figure 2A schematic diagram of the component structure of a cold chain logistics service system provided in an embodiment of the present application for implementing the above-mentioned cold chain transportation demand analysis method based on big data prediction is shown. DETAILED DESCRIPTION

[0017] The embodiments of the present application are described below in conjunction with the drawings in the present application. It should be understood that the implementation methods described below in conjunction with the drawings are exemplary descriptions for explaining the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions of the embodiments of the present application.

[0018] It will be understood by those skilled in the art that, unless specifically stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the terms "including" and "comprising" used in the embodiments of the present application refer to that the corresponding features can be implemented as the presented features, information, data, steps, operations, elements and / or components, but do not exclude the implementation as other features, information, data, steps, operations, elements, components and / or combinations thereof supported by the technical field. It should be understood that when an element is "connected" or "coupled" to another element, the element may be directly connected or coupled to the other element, or may refer to the element and the other element establishing a connection relationship through an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling, and the term "and / or" used herein indicates at least one of the items defined by the term, such as "A and / or B" may be implemented as "A", or as "B", or as "A and B".

[0019] In order to make the purpose, technical solution and advantages of the present application clearer, the implementation mode of the present application will be further described in detail with reference to the accompanying drawings. The technical solution of the embodiment of the present application and the technical effect produced by the technical solution of the present application are explained below by describing several exemplary implementation modes. It should be pointed out that the following implementation modes can refer to, draw lessons from or combine with each other, and the same terms, similar features and similar implementation steps in different implementation modes are not described repeatedly.

[0020] Figure 1 The flowchart of the cold chain transportation demand analysis method and system based on big data prediction provided by the embodiment of the present application is shown. It should be understood that in other embodiments, the order of some steps in the cold chain transportation demand analysis method based on big data prediction of this embodiment can be shared with each other according to actual needs, or some steps can be omitted or maintained. The detailed steps of the cold chain transportation demand analysis method based on big data prediction include: Step S110, obtaining a historical cold chain order dataset of a target area, wherein the historical cold chain order dataset includes temperature-sensitive cargo transportation records and corresponding environmental parameters for multiple time periods.

[0021] In this embodiment, in the fresh food scenario, the cold chain transportation of fruits and beef is taken as an example. Assume that the target area is a large fruit and beef supply area, such as the main agricultural production area and livestock breeding area of ​​a province. For fruits, it may include apples, bananas, oranges and other varieties, and beef comes from major local farms. Multiple time periods can be every month or every quarter in the past year.

[0022] In terms of fruit transportation records, the transportation records of apples show that in spring, because apples have been stored for a long time and have just been taken out of cold storage, the temperature requirements are relatively low, and the temperature during transportation is kept between 2-4°C. Bananas are more special. When transported in the summer ripening season, the temperature needs to be strictly controlled at 12-14°C, because bananas tend to mature and rot quickly under high temperatures, while the temperature can be appropriately lowered to 10-12°C during winter transportation. It is best to keep the temperature of oranges at 5-7°C during transportation. At the same time, the records also include the origin and destination of the transportation, such as apples transported from orchards to fruit wholesale markets, or from wholesale markets to various supermarkets.

[0023] For beef, its transportation records show that the temperature of fresh beef needs to be kept between -2-0℃ during transportation to ensure the freshness of the meat and prevent bacteria from growing. When the transportation distance is long, such as from the breeding area to a distant city, stricter temperature control and faster transportation time are required.

[0024] In terms of the corresponding environmental parameters, during the transportation process, the external temperature in different seasons has a great impact on cold chain transportation. In summer, the external temperature is often as high as 30°C or above, which puts higher requirements on refrigeration equipment and requires more energy to maintain the required temperature of the goods. In winter, especially in cold areas, the external temperature may be as low as -10°C or below. Although the energy consumption of refrigeration equipment will be reduced, it is also necessary to prevent the goods from being frostbitten. In addition, humidity is also an important environmental parameter. A high humidity environment may cause fruits to mold and rot, and for beef, it may affect the dryness and shelf life of its surface. Road conditions are also part of the environmental parameters. For example, the roads in mountainous areas are rugged, and the bumps during transportation may affect the normal operation of refrigeration equipment and may also affect the quality of fruits and beef.

[0025] Step S120, extracting a set of transportation demand features from the historical cold chain order data, wherein the set of transportation demand features includes correlation features between cargo types and temperature sensitivity, balance features between transportation timeliness and equipment energy consumption, and matching features between regional climate and storage conditions.

[0026] As for the correlation characteristics between cargo types and temperature sensitivity, we still take fruits and beef as examples. Bananas among fruits are extremely temperature-sensitive cargoes. As mentioned earlier, slight fluctuations in temperature may affect their ripening speed and quality. When the temperature is higher than 14°C, bananas may begin to turn yellow and black quickly during transportation, affecting sales. Apples are relatively less sensitive to temperature, but if the temperature is too high, they will also lose water and deteriorate the taste. As a meat product, beef is also highly sensitive to temperature, and the temperature range of -2-0°C is the key to ensuring its freshness. If the temperature rises, the bacteria will multiply faster, causing the beef to deteriorate.

[0027] In terms of the balance between transportation time and equipment energy consumption, consider the transportation of fruits and beef. Suppose a batch of fresh bananas are transported from the place of production to a distant big city, which is a long distance. If the bananas are to arrive at the destination in the best quality state, a faster transportation time is required. This means that the transportation vehicles need to maintain a high speed, and the refrigeration equipment needs to operate continuously and stably, consuming a lot of energy. However, if the transportation time is too long, even if the refrigeration equipment operates normally, the bananas may deteriorate due to their own physiological characteristics. For beef, transportation time is equally important to ensure the freshness of the meat. If a slower transportation method is used, although the energy consumption of the refrigeration equipment will be reduced, the beef may exceed the shelf life during transportation. For example, if the transportation time exceeds a certain limit when transporting beef from the breeding area to a distant processing enterprise, even if the temperature is kept at -2-0℃, the freshness of the beef will be affected, the color may become darker, and the microbial indicators may exceed the standard.

[0028] The matching characteristics of regional climate and storage conditions are also very critical. In the hot and humid southern region, the storage conditions of fruits and beef require stronger refrigeration capacity and dehumidification functions. For example, in southern cities in summer, fruit warehouses need to maintain low temperatures and humidity to prevent the fruits from mold and rot. For beef warehouses, not only the temperature must be controlled, but also air circulation must be ensured to prevent odors. In the cold and dry northern region, storage conditions need to consider the problem of preventing goods from being frostbitten. For example, in the north in winter, fruit warehouses need appropriate insulation measures to prevent the fruits from being frozen, and beef warehouses also need to adjust the temperature control range to prevent the beef from affecting the taste due to excessive freezing.

[0029] Step S130, input the transport demand feature set into a pre-trained spatiotemporal prediction model to generate multi-dimensional cold chain demand parameters for the target area within a future specified period, wherein the multi-dimensional cold chain demand parameters include a temperature fluctuation tolerance threshold, a transport timeliness critical value, and a path node load coefficient.

[0030] Taking the cold chain transportation of fruits and beef as an example, the previously extracted set of transportation demand features is input into the spatiotemporal prediction model within a specified period in the future (such as the upcoming summer).

[0031] As for the temperature fluctuation tolerance threshold, due to the high temperature in summer, the temperature fluctuation tolerance threshold of bananas among fruits may be very small, for example, it can only fluctuate ±0.5℃ between 12-14℃, because bananas are easy to deteriorate under high temperature. The temperature fluctuation tolerance threshold of apples may be relatively large, fluctuating ±1℃ between 2-4℃. The temperature fluctuation tolerance threshold of beef fluctuates ±0.3℃ between -2-0℃, because beef is very sensitive to small changes in temperature.

[0032] In terms of the critical value of transportation time, in the summer fruit harvest season, the critical value of transportation time for fruits is shorter. For example, if bananas cannot arrive from the production area to the market within 3 days, they may lose their best sales quality due to temperature and physiological changes. The critical value of transportation time for apples may be around 5 days. The critical value of transportation time for beef from farms to processing companies or sales markets may be shorter, such as 2 days, because freshness is crucial to the quality and sales of beef.

[0033] In terms of the path node load factor, it is assumed that there are multiple fruit and beef transportation nodes in the target area, such as the collection center of the origin, the transfer station during transportation, the wholesale market at the destination, etc. In summer, due to the large transportation volume of fruit and beef, the load factors of these nodes will increase accordingly. For example, a collection center in a fruit production area has to handle a large number of fruit transportation tasks every day during the summer fruit harvest period, and its load factor may reach 0.8, indicating that it is close to saturation. In winter, the load factor may be only 0.3. For nodes for beef transportation, such as collection points near farms, the load factor will also increase during peak beef demand periods such as holidays, approaching full load.

[0034] Step S140, generating a cold chain transportation dynamic scheduling plan with a timestamp based on the multi-dimensional cold chain demand parameters.

[0035] Based on the multi-dimensional cold chain demand parameters obtained above, a dynamic scheduling plan for cold chain transportation of fruits and beef is formulated.

[0036] For fruit transportation, take bananas as an example. Due to its low tolerance threshold for temperature fluctuations and short critical value for transportation time, it is necessary to accurately arrange the transportation time and route during transportation. Assuming that bananas are transported from origin A to the fruit market in city B, according to the temperature fluctuation tolerance threshold, select a transportation vehicle with good refrigeration effect and precise temperature control. In summer, the transportation time must be strictly controlled within 3 days. In terms of transportation route planning, avoid high-temperature sections or traffic congestion sections to reduce transportation time. In terms of departure time, choose to depart in the early morning when the temperature is lower, and during transportation, dynamically adjust the power of the refrigeration equipment according to real-time temperature data. For example, when passing through a high-temperature area, increase the refrigeration power in advance to ensure that the temperature in the car is always maintained between 12-14°C.

[0037] For beef transportation, because its temperature fluctuation tolerance threshold is small and the critical value of transportation time is short, the transportation vehicles from farm C to the processing enterprise in city D must have good insulation and refrigeration capabilities. The departure time must ensure that the beef reaches the destination in the shortest time, such as departing at night when the outside temperature is relatively low, which helps to reduce the energy consumption of refrigeration equipment. The transportation route should choose roads with good road conditions and short distances, and during the transportation process, check the temperature in the car and the state of the beef at regular intervals. If the temperature is found to be rising and approaching the temperature fluctuation tolerance threshold, adjust the refrigeration equipment in time.

[0038] In terms of path node load, for each node of fruit and beef transportation, such as collection centers, transfer stations, etc., the entry and exit of goods are arranged according to the load factor. During the summer fruit harvest period, if the load factor of the collection center is close to saturation, priority will be given to the transportation of perishable fruits such as bananas to ensure that they can leave the collection center as soon as possible and reduce the residence time. For the nodes of beef transportation, during the peak demand period, the order of vehicle loading and unloading is reasonably arranged to improve the turnover efficiency of the nodes. At the same time, according to the demand and node load conditions in different time periods, a specific timestamp is assigned to each transportation task, and the start time, arrival time and residence time at each node of the transportation are accurately arranged to ensure that the quality of fruits and beef is not affected during the entire cold chain transportation process.

[0039] Based on the above steps, the embodiment of the present application obtains a historical cold chain order data set of temperature-sensitive cargo transportation records and their corresponding environmental parameters in the target area for multiple time periods, associates the transportation records with the environmental parameters, and defines a set of transportation demand characteristics, among which the association characteristics of cargo types and temperature sensitivity break the previous simple classification of cargo temperature requirements, deeply explore the intrinsic connection between different cargoes in temperature sensitivity, and provide a basis for more accurate control of transportation temperature; the balance characteristics of transportation timeliness and equipment energy consumption solve the long-standing problem of balancing timeliness and energy consumption in cold chain transportation, and provide a new perspective for optimizing transportation strategies by analyzing the relationship between the two; the matching characteristics of regional climate and storage conditions combine the external environment with the internal storage conditions, which is a key connection overlooked in previous cold chain transportation demand analysis and helps to rationally plan storage layout and condition settings.

[0040] In the prediction stage, the extracted transport demand feature set is input into the pre-trained spatiotemporal prediction model to generate multi-dimensional cold chain demand parameters for the target area within the specified future period. The generation of multi-dimensional parameters such as temperature fluctuation tolerance threshold, transport timeliness critical value and path node load coefficient can comprehensively and forward-lookingly characterize the cold chain transport demand compared to the traditional single-dimensional prediction, allowing transport companies to understand in advance the future transportation process in terms of temperature control, time scheduling, resource allocation and other requirements.

[0041] Traditional scheduling solutions often lack comprehensive consideration of time dimensions and multi-dimensional demand parameters, and are difficult to adapt to the complex and changing cold chain transportation environment. However, this solution, by combining timestamps and multi-dimensional parameters, can scientifically schedule transportation needs at different times in the future in real time and dynamically, thereby improving the utilization efficiency of cold chain transportation resources, reducing transportation costs, and maximizing the transportation quality of temperature-sensitive goods, thus improving the intelligence and adaptability of the entire cold chain transportation system.

[0042] In a possible implementation, step S140 includes: Step S141, based on the multi-dimensional cold chain demand parameter optimization, a cold chain transportation path topology is generated, and each transportation node in the cold chain transportation path topology is associated with a dynamic resource allocation weight, and the dynamic resource allocation weight is negatively correlated with the historical failure rate of the corresponding node, the real-time temperature control capability, and the connectivity efficiency of adjacent nodes.

[0043] In this embodiment, the cold chain transportation of fruits and beef is used as an example to explain this process in detail. For the transportation of bananas among fruits, consider the transportation link from the banana plantation to the fruit market in the distant city. The nodes on the transportation path include the collection center near the plantation, the transfer station during transportation, and the wholesale warehouse in the destination city. If the historical failure rate of the collection center is high, for example, there have been multiple refrigeration equipment failures in the past year due to equipment aging and other reasons, resulting in large temperature fluctuations in the collection process of bananas, affecting the quality, then its dynamic resource allocation weight will be low. Real-time temperature control capability is also an important factor. If the current refrigeration equipment of the collection center can only control the temperature at 13-15°C, and the optimal transportation temperature of bananas is 12-14°C, which exceeds the suitable temperature range, this indicates that its temperature control capability is insufficient, which will also reduce the dynamic resource allocation weight. In terms of the connectivity efficiency of adjacent nodes, if the road conditions between the collection center and the next transfer station are poor, traffic congestion often occurs, resulting in extended transportation time and low connectivity efficiency, which will also reduce the dynamic resource allocation weight of the collection center.

[0044] For beef transportation, on the path from the farm to the processing enterprise or sales market, if the collection point near the farm has historically had refrigeration equipment failures due to unstable power supply, affecting the preservation of beef, its dynamic resource allocation weight will be reduced. If the real-time temperature control capability of the collection point can only maintain the temperature of the beef at -1-1°C, while the ideal temperature is -2-0°C, this will affect the freshness of the beef and reduce its dynamic resource allocation weight. In terms of the connectivity efficiency of adjacent nodes, if the transportation route between the collection point and the next transportation node is often affected by the weather, such as snow and ice accumulation in winter, which affects the transportation speed and cargo safety, this will reduce its dynamic resource allocation weight. Based on these factors, the cold chain transportation path topology is generated through the optimization algorithm, and nodes with higher dynamic resource allocation weights are preferentially selected to form a transportation path to ensure the quality and efficiency of fruit and beef during transportation.

[0045] Step S142, according to the real-time operating status data of each node in the cold chain transportation path topology, dynamically adjust the matching relationship between the transportation time limit critical value and the temperature fluctuation tolerance threshold, and generate an abnormal event response strategy set, wherein the abnormal event response strategy set includes the activation conditions of the standby refrigeration equipment, the path switching priority rules and the emergency storage capacity allocation plan.

[0046] Let's continue with the cold chain transportation of fruits and beef. In the transportation of fruits, for example, when bananas are being transported from the collection center to the transfer station, real-time monitoring shows that the refrigeration equipment of the transport vehicle has a minor fault, causing the temperature inside the vehicle to rise from the normal range of 12-14°C to 14-15°C, close to the upper limit of the temperature fluctuation tolerance threshold of bananas. At this time, according to the real-time operating status data of each node in the transportation path topology, since the failure of the refrigeration equipment affects the temperature control during transportation, it is necessary to dynamically adjust the matching relationship between the transportation time limit and the temperature fluctuation tolerance threshold. The original transportation time limit for bananas from the collection center to the transfer station is 1 day, but due to the temperature rise, in order to ensure the quality of bananas, the transportation time limit may be shortened to half a day. At the same time, the condition for activating the standby refrigeration equipment is set to start the standby refrigeration equipment immediately when the temperature rises to 15°C to reduce the temperature inside the vehicle.

[0047] For the path switching priority rule, if there is severe traffic congestion on the current transportation route, which leads to a significant extension of the transportation time and threatens the quality of the bananas, then the backup transportation route will be switched first. In terms of the emergency storage capacity allocation plan, if some bananas are found to have shown signs of slight deterioration due to temperature fluctuations when they arrive at the transfer station, then according to the emergency storage capacity allocation plan, bananas with a higher risk of deterioration will be allocated to the emergency storage area first. The temperature and humidity conditions in this area are more suitable for temporary storage and inspection of these bananas to prevent further deterioration of the deterioration.

[0048] In terms of beef transportation, suppose that when beef is transported from the collection point of the farm to the processing enterprise, the real-time monitoring detects a sudden rise in the external temperature. Although the refrigeration equipment is operating normally, the temperature inside the vehicle rises from the normal range of -2-0℃ to -1-0.5℃, which is close to the upper limit of the temperature fluctuation tolerance threshold of beef. At this time, the critical value of transportation time is shortened, and the original 2-day transportation time may be adjusted to 1.5 days. The activation condition of the standby refrigeration equipment is set to start when the temperature rises to 0.5℃. If force majeure factors such as natural disasters on the transportation route make the road impassable, according to the path switching priority rule, priority is given to alternative routes with better road conditions, relatively short distances and able to ensure the temperature requirements of beef. When the beef arrives at the storage point near the processing enterprise, if the color and smell of some beef are found to be abnormal, according to the emergency storage capacity allocation plan, this part of the beef will be stored separately in the emergency storage area. The temperature and ventilation conditions in this area are more conducive to further inspection and processing of the beef to prevent spoiled beef from affecting other normal beef.

[0049] Step S143, perform spatiotemporal alignment processing on the abnormal event response strategy set and the real-time meteorological data stream to generate a dynamic scheduling plan for cold chain transportation with a timestamp, wherein the dynamic scheduling plan for cold chain transportation includes a pre-start time window for equipment, a transport batch interval threshold, and a trigger condition for cross-regional collaborative transportation.

[0050] When performing spatiotemporal alignment processing with real-time meteorological data streams, take banana transportation as an example. If the meteorological data predicts that high temperatures will occur on the transportation route in the next few hours, then according to the abnormal event response strategy set, the equipment pre-start time window will be opened in advance. For example, under normal circumstances, the refrigeration equipment is started 15 minutes before loading. Now, due to the high temperature forecast, the pre-start time window is advanced to 30 minutes before loading to ensure that the temperature in the vehicle is within the suitable transportation temperature range of 12-14°C for bananas when loading. In terms of the transportation batch interval threshold, if the meteorological data shows that the weather conditions on the transportation route are unstable within a certain period of time, in order to avoid different batches of bananas affecting each other during transportation, for example, high temperature may increase the burden on the refrigeration equipment, and if the batch interval is too small, it may affect the refrigeration effect, then the transportation batch interval threshold is increased, and the original batch of goods was sent every 2 hours, and now it is adjusted to send a batch of goods every 3 hours.

[0051] When it comes to cross-regional coordinated transportation trigger conditions, if the banana production in a certain area suddenly increases significantly, and the local market has limited digestion capacity, and at the same time, meteorological data shows that the surrounding areas have suitable meteorological conditions for banana transportation, and there is demand for bananas in the surrounding regional markets, then cross-regional coordinated transportation is triggered. For example, the production in the area where the local banana plantation is located has increased significantly, and there is demand in the urban fruit market in the surrounding areas. Meteorological data shows that the weather on the transportation route is good. At this time, cross-regional coordinated transportation is initiated to transport bananas to the surrounding areas, which not only solves the sales problem of local bananas, but also meets the market demand of the surrounding areas.

[0052] For beef transportation, meteorological data predicts that low temperatures will occur during transportation, which may cause frostbite of beef. According to the set of abnormal event response strategies, the equipment pre-start time window is adjusted to preheat the insulation equipment of the transport vehicle in advance to ensure that the temperature in the vehicle is in the appropriate range of -2-0℃ when loading. In terms of the transport batch interval threshold, if the meteorological data shows that severe weather such as heavy snow may occur on the transportation route within a certain period of time, in order to avoid the impact of different batches of beef during transportation, the transport batch interval threshold is increased. For example, the original batch of goods was sent every 1.5 hours, and now it is adjusted to send a batch every 2.5 hours. In terms of cross-regional collaborative transportation trigger conditions, if a farm has an excess of beef production and the processing capacity of local processing companies is limited, and the meteorological data shows that other regions have suitable transportation and processing conditions and there is demand for beef, then cross-regional collaborative transportation is triggered to transport beef to other regions for processing or sales, so as to achieve effective allocation of resources. In this way, a dynamic scheduling plan for cold chain transportation with timestamps is generated to accurately arrange each link of the cold chain transportation process of fruits and beef to ensure the quality and transportation efficiency of goods.

[0053] In a possible implementation, step S120 includes: Step S121, identifying the correspondence between the cargo category codes in the historical cold chain order data and the temperature monitoring curve, and generating a temperature sensitivity distribution map.

[0054] In this embodiment, taking the cold chain transportation of fruits and beef as an example, in the historical cold chain order data, for bananas among fruits, its cargo category code corresponds to a specific temperature monitoring curve. The cargo category code of bananas is a specific identifier, such as "F-001", which uniquely identifies bananas as a cargo in the entire cold chain transportation data system. In terms of the temperature monitoring curve, starting from the time when bananas are picked, the temperature is strictly controlled at around 13-14°C during storage, and this temperature range can delay the ripening process of bananas to the greatest extent. When bananas are loaded into the transport vehicle, the temperature monitoring device begins to continuously record temperature changes. If the transportation distance is long, the temperature may rise slightly at the beginning of transportation, for example, from 13.5°C to 14°C, because of the heat exchange during vehicle startup and cargo loading. With the stable operation of the refrigeration equipment, the temperature will gradually stabilize between 13-14°C. Before arriving at the destination, the outside temperature may be slightly higher due to the proximity of the vehicle to the urban area, and the temperature will fluctuate slightly again.

[0055] For apples, the cargo category code is "F-002", and its temperature monitoring curve is different from that of bananas. The temperature of apples can be stored at 2-4℃. After picking, they enter the warehouse and the temperature is kept within this range. During transportation, the temperature fluctuation is relatively small, basically maintained at 2-3℃. This is because apples are slightly less temperature sensitive than bananas. The cargo category code of beef is "M-001", and its temperature monitoring curve shows that after slaughter, beef is quickly cooled to -2-0℃. During transportation, this temperature range must be strictly maintained. During transportation from the farm to the processing plant or sales market, the temperature curve recorded by the temperature monitoring equipment fluctuates very little. Once the temperature exceeds the range of -2-0℃, the freshness of the beef will be affected and the risk of bacterial growth will increase significantly. By identifying the corresponding relationship between these different cargo category codes and temperature monitoring curves, a temperature sensitivity distribution map is generated. This temperature sensitivity distribution map intuitively shows the temperature sensitivity characteristics of different goods during cold chain transportation, providing an important basis for subsequent cold chain transportation scheduling.

[0056] Step S122, analyzing the correlation pattern between the transport time efficiency record and the refrigeration equipment energy consumption log, and establishing a time efficiency-energy consumption balance coefficient matrix.

[0057] Taking banana transportation as an example, when the transportation distance is 500 kilometers and the transportation time is required to reach the destination within 3 days, the refrigeration equipment needs to run continuously to maintain a temperature of 13-14°C. The refrigeration equipment energy consumption log shows that during this transportation process, the energy consumption of the refrigeration equipment shows a certain change pattern with the passage of transportation time. In the early stage of transportation, since the vehicle needs to reduce the external ambient temperature to a suitable transportation temperature for bananas, the refrigeration equipment needs to run at high power and consumes high energy. When the temperature stabilizes, the energy consumption will decrease, but energy still needs to be consumed continuously to maintain the temperature. If the transportation time is shortened to 2 days, in order to ensure the quality of the bananas, the vehicle needs to increase the driving speed, which may cause the refrigeration equipment to be subject to more vibration and external environmental interference. The refrigeration equipment needs to adjust the refrigeration power more frequently, resulting in increased energy consumption.

[0058] For apple transportation, assume that the transportation distance is 800 kilometers and the transportation time is 5 days. After the refrigeration equipment reduces the temperature to 2-4℃ at the beginning of transportation, since the temperature sensitivity of apples is relatively low compared with bananas, the refrigeration equipment does not need to adjust the power frequently, and the energy consumption is relatively stable. However, if the transportation time is shortened to 3 days, in order to meet the faster transportation speed, the refrigeration equipment may need to adjust the refrigeration power more flexibly during the acceleration and deceleration of the vehicle to cope with the impact of changes in the external environment on the temperature of the compartment, which will also lead to increased energy consumption. In beef transportation, the transportation distance is 300 kilometers and the transportation time is 2 days. Since beef has strict temperature requirements, the temperature range of -2-0℃ must always be maintained. The energy consumption of refrigeration equipment is relatively stable during transportation, but if the transportation time is shortened to 1.5 days, the vehicle speed is increased, and the refrigeration equipment needs to control the temperature more accurately to prevent the temperature of beef from fluctuating, which may require higher refrigeration power, thereby increasing energy consumption. By analyzing the energy consumption of refrigeration equipment for these different goods under different transportation times, a time-efficiency-energy consumption balance coefficient matrix is ​​established. The elements in this matrix represent the quantitative relationship between the energy consumption of refrigeration equipment required to maintain the temperature of goods under different transportation time limits, providing key data for cost control and efficiency optimization of cold chain transportation.

[0059] Step S123, detecting the matching degree between the regional climate data and the storage temperature maintenance capacity, and calculating the climate adaptability index.

[0060] Take banana storage as an example. In tropical and subtropical areas in the south, the climate is hot and humid, with summer temperatures often reaching over 30°C and high humidity. Banana storage facilities need to have strong refrigeration and dehumidification capabilities. If the storage temperature maintenance capacity is insufficient, for example, the refrigeration equipment can only reduce the temperature to 15-16°C, and cannot reach the optimal storage temperature of 13-14°C for bananas, then the storage quality of bananas will be affected, and the bananas may mature and rot too quickly. At this time, the regional climate data is less compatible with the storage temperature maintenance capacity.

[0061] For apple storage, in the temperate zone of the north, the winter is cold, the temperature may drop to below -10℃, and the summer is relatively mild. Apple storage facilities need to prevent apples from being frostbitten in winter and maintain a suitable temperature in summer. If the storage temperature maintenance capacity can be flexibly adjusted according to seasonal changes, for example, the refrigeration can be appropriately reduced in winter, and the temperature can be stably maintained at 2-4℃ in summer, then the regional climate data and the storage temperature maintenance capacity are highly matched. For beef storage, whether in the cold north or the hot south, the temperature of -2-0℃ needs to be strictly maintained. In the hot and humid southern region, storage facilities need better insulation and refrigeration capabilities to combat the external climate. If the storage temperature maintenance capacity can effectively cope with the external climate and ensure the freshness of the beef, then the regional climate data and the storage temperature maintenance capacity are highly matched. By testing the climate data and storage temperature maintenance capacity of different regions, the climate adaptability index is calculated. This index quantifies the matching relationship between the regional climate and the storage temperature maintenance capacity, and provides an important reference for the storage site selection and facility configuration during cold chain transportation.

[0062] Step S124, integrating the temperature sensitivity distribution map, the time efficiency-energy consumption balance coefficient matrix and the climate adaptability index to generate a three-dimensional feature vector space as the transportation demand feature set.

[0063] For example, the temperature sensitivity distribution map describes the temperature sensitivity characteristics of goods such as bananas, apples and beef during transportation, which is one of the core requirements of cold chain transportation. The time-energy balance coefficient matrix reflects the energy consumption relationship of refrigeration equipment required to ensure the temperature of goods under different transportation time, which involves the cost and efficiency of cold chain transportation. The climate adaptability index quantifies the matching relationship between regional climate and storage temperature maintenance capacity, which is of great significance to the storage link of cold chain transportation.

[0064] Taking banana transportation as an example, the temperature sensitivity distribution map shows that bananas can only maintain good quality within a temperature range of 13-14°C. This temperature requirement occupies one dimension in the three-dimensional feature vector space. The time-energy balance coefficient matrix shows the energy consumption of refrigeration equipment when the transportation time is 3 days and the transportation distance is 500 kilometers. This data occupies another dimension in the three-dimensional feature vector space. The climate adaptability index reflects the adaptability of banana storage facilities to local climate conditions. For example, in the hot and humid southern region, this index is lower, and this data also becomes a dimension of the three-dimensional feature vector space.

[0065] For apples and beef, their respective temperature sensitivity, time-efficiency-energy consumption relationship, and climate adaptability data are also integrated into this three-dimensional feature vector space. This three-dimensional feature vector space comprehensively covers the key transportation demand characteristics of fruits and beef in the cold chain transportation process. As a collection of transportation demand characteristics, it provides comprehensive and accurate data input for the subsequent spatiotemporal prediction model, which helps to more accurately predict the demand parameters of cold chain transportation, thereby optimizing the scheduling and management of cold chain transportation.

[0066] The training step of the spatiotemporal prediction model includes: Step S101: divide the three-dimensional feature vector space into a training set and a validation set, and annotate the training set and the validation set with corresponding time series labels.

[0067] In the cold chain transportation scenario of fruit and beef, the generated three-dimensional feature vector space is divided into training set and validation set. Assume that the three-dimensional feature vector space contains the cold chain transportation data of fruit and beef in the past year. According to a certain ratio, for example, 80% of the data is divided into training set and 20% of the data is divided into validation set.

[0068] For banana transportation data, the training set includes data on temperature sensitivity, time-efficiency-energy consumption balance coefficient, and climate adaptability under different seasons, different transportation distances, and different transportation time. For example, a set of data on banana transportation in spring with a transportation distance of 300 kilometers and a transportation time of 2 days has a corresponding time series label of spring, 300 kilometers, and 2 days. This label accurately describes the time and transportation-related characteristics of this set of data. The validation set also includes similar banana transportation data, but it is different from the training set data. For example, a set of data on banana transportation in summer with a transportation distance of 400 kilometers and a transportation time of 2.5 days has a time series label of summer, 400 kilometers, and 2.5 days.

[0069] For the transportation data of apples and beef, the training set and validation set are also divided in this way, and the corresponding time series labels are annotated. The apple transportation data may contain autumn transportation in the training set, with a transportation distance of 600 kilometers and a transportation time of 4 days, and the labels are autumn, 600 kilometers, and 4 days. In the validation set, there may be winter transportation, with a transportation distance of 500 kilometers and a transportation time of 3.5 days, and the labels are winter, 500 kilometers, and 3.5 days. In the training set, the beef transportation data may contain data from a farm to a specific processing enterprise, with a transportation distance of 200 kilometers and a transportation time of 1.5 days, and the labels are specific time, 200 kilometers, and 1.5 days. In the validation set, there may be data from another farm to a different processing enterprise, with a transportation distance of 250 kilometers and a transportation time of 1.8 days, and the labels are specific time, 250 kilometers, and 1.8 days. In this way, the three-dimensional feature vector space is accurately divided into training sets and validation sets, and the corresponding time series labels are annotated for them, providing an organized and labeled data source for the subsequent spatiotemporal prediction model training.

[0070] Step S102, constructing a deep spatiotemporal convolutional network, wherein the deep spatiotemporal convolutional network includes a climate fluctuation perception layer, a transportation path dependency layer, and an equipment state memory unit.

[0071] For example, the climate fluctuation perception layer is mainly responsible for processing the impact of regional climate data on cold chain transportation. Taking banana transportation as an example, when bananas are transported from tropical areas to temperate areas, the climate fluctuation perception layer will receive high temperature and high humidity climate data from tropical areas and relatively mild climate data from temperate areas. It can analyze the impact of climate change from tropical to temperate zones on temperature control of banana transportation during transportation. For example, from a high temperature environment to a relatively low temperature environment, refrigeration equipment may need to make appropriate power adjustments during transportation to prevent bananas from being affected by sudden temperature changes.

[0072] The transport path dependency layer focuses on factors related to the transport path. For apple transportation, the transport path from the orchard to the wholesale market may pass through mountainous areas and plains. The transport path dependency layer will consider the impact of rugged mountain roads on transportation speed, thereby affecting transportation timeliness, as well as the impact of traffic flow in plains on transportation. If there are often congested sections on the transport path, this layer will take it into consideration, because congestion will extend the transportation time, which in turn affects the energy consumption of refrigeration equipment and the preservation of apples.

[0073] The device state memory unit is used to record the status information of cold chain transportation equipment such as refrigeration equipment. In beef transportation, the device state memory unit will record the working hours of the refrigeration equipment, changes in refrigeration efficiency, and whether there have been any failures. For example, the refrigeration efficiency of the refrigeration equipment may decrease after long-term operation. The device state memory unit will record this change and feed it back to the entire deep spatiotemporal convolutional network. This equipment state information will be taken into account when predicting the temperature control requirements during beef transportation. By constructing a deep spatiotemporal convolutional network that includes a climate fluctuation perception layer, a transportation path dependency layer, and a device state memory unit, various factors in the cold chain transportation process of fruits and beef can be fully considered, so as to more accurately predict demand parameters.

[0074] Step S103, optimizing the network parameters of the deep spatiotemporal convolutional network by a back-propagation algorithm to minimize the spatiotemporal correlation loss function between the predicted output of the deep spatiotemporal convolutional network and the actual required parameters of the validation set.

[0075] For example, taking banana transportation as an example, after inputting the training set data, the deep spatiotemporal convolutional network will output the predicted values ​​of the demand parameters for banana transportation, such as the temperature fluctuation tolerance threshold, the critical value of transportation time, etc. These predicted values ​​are then compared with the actual demand parameters of the validation set. Suppose there is a set of banana transportation data in the validation set, the actual temperature fluctuation tolerance threshold is ±0.5℃, and the critical value of transportation time is 3 days. The initial prediction value of the deep spatiotemporal convolutional network is the temperature fluctuation tolerance threshold of ±1℃, and the critical value of transportation time is 3.5 days.

[0076] Calculate the spatiotemporal correlation loss function, which comprehensively considers the prediction errors of the temperature fluctuation tolerance threshold and the transportation timeliness critical value. Since there is a difference between the predicted value and the true value, the value of the loss function is large. Through the back propagation algorithm, the value of the loss function is back-propagated from the output layer to each layer of the network to adjust the network parameters, such as the adjustment of the climate data weight in the climate fluctuation perception layer, the adjustment of the path factor weight in the transportation path dependency layer, and the adjustment of the equipment status information weight in the equipment status memory unit.

[0077] The same process is used for the transportation of apples and beef. For example, in the transportation of apples, the actual demand parameters of the validation set are the temperature fluctuation tolerance threshold of ±1°C and the transportation time limit of 5 days, while the network's initial prediction value is the temperature fluctuation tolerance threshold of ±1.5°C and the transportation time limit of 5.5 days. The network parameters are adjusted through the back-propagation algorithm to gradually reduce the loss function. In the transportation of beef, assuming that the actual demand parameters of the validation set are the temperature fluctuation tolerance threshold of ±0.3°C and the transportation time limit of 2 days, the network's initial prediction value is the temperature fluctuation tolerance threshold of ±0.5°C and the transportation time limit of 2.2 days. By continuously back-propagating and adjusting the network parameters, the spatiotemporal correlation loss function between the prediction output of the deep spatiotemporal convolutional network and the actual demand parameters of the validation set is minimized, thereby improving the prediction accuracy of the deep spatiotemporal convolutional network for the cold chain transportation demand parameters of fruits and beef.

[0078] Step S104, when the loss value change rate of N consecutive training cycles is lower than a preset threshold, the network parameters of the deep spatiotemporal convolutional network are frozen and the model weights are derived.

[0079] For example, suppose the preset threshold is 0.01 and N is 5. During the training process, the loss value change rate is calculated for each training cycle. Taking banana transportation as an example, in the first few training cycles, as the back propagation algorithm adjusts the network parameters, the loss value will gradually decrease and the loss value change rate will also be large. However, as the training progresses, when the loss value change rate for 5 consecutive training cycles is less than 0.01, it means that the deep spatiotemporal convolutional network has converged to a better state.

[0080] The same is true for the transportation of apples and beef. For example, in the deep spatiotemporal convolutional network training related to apple transportation, when the loss value change rate for 5 consecutive training cycles is less than 0.01, it indicates that the network's prediction of the cold chain transportation demand parameters of apples has reached a relatively stable and accurate state. Similarly, in the deep spatiotemporal convolutional network training related to beef transportation, when this condition is met, it means that the network's prediction ability for the cold chain transportation demand parameters of beef has been optimized to a certain extent. At this time, the network parameters of the deep spatiotemporal convolutional network are frozen and no longer adjusted, and then the model weights are derived. These model weights contain the parameter information of each layer such as the optimized climate fluctuation perception layer, transportation path dependency layer, and equipment state memory unit. These weights can be used for subsequent cold chain transportation demand parameter predictions, providing an accurate basis for the scheduling and management of cold chain transportation.

[0081] In a possible implementation, step S141 includes: Step S1411, calculating the bearing margin value of each transport section according to the path node load coefficient.

[0082] In this embodiment, in the cold chain transportation scenario of fruit and beef, the transportation path from the fruit production area to the urban sales market and from the beef farm to the processing enterprise or sales point is considered. The path contains multiple nodes, such as the collection center of the fruit production area, the transfer station during transportation, the distribution center on the edge of the city, the collection point of the farm in beef transportation, the refrigerated warehouse in the middle, and the processing workshop at the destination. Taking the transportation section from the collection center to the transfer station in fruit transportation as an example, the path node load factor reflects the busyness and resource occupancy of the node. Assuming that the collection center has to handle the collection and temporary storage of 100 boxes of fruit every day in a certain period of time, and its designed maximum processing capacity is 150 boxes, then its load factor is 100 / 150=0.67. The design receiving capacity of the transfer station is 200 boxes per day, the actual receiving volume is 120 boxes, and the load factor is 120 / 200=0.6. For this transport section, the carrying margin value is equal to 1-(0.67+0.6)=-0.27 (this is only an example calculation, and it may be adjusted according to the specific calculation logic and data in practice, and factors such as transportation direction and node weight may need to be considered). In the case of beef transportation, from the collection point of the farm to the intermediate cold storage warehouse, assuming that the collection point handles the collection of 50 cattle per day, the maximum processing capacity is 80, and the load factor is 50 / 80=0.625, and the cold storage warehouse receives 40 cattle per day, the maximum receiving capacity is 60, and the load factor is 40 / 60=0.667. The carrying margin value of this transport section is 1-(0.625+0.667)=-0.292. Through such calculations, the carrying margin value of each transport section can be obtained. This value reflects the remaining carrying capacity of the transport section and provides basic data for the subsequent construction of the cold chain transportation path topology.

[0083] Step S1412, constructing a topological relationship graph between transportation nodes based on a graph neural network, wherein the edge weights in the topological relationship graph are determined by the product of the carrying margin value and the transportation time efficiency.

[0084] Still taking the cold chain transportation of fruit and beef as an example, for the nodes such as the collection center, transfer station, and distribution center on the fruit transportation path, and the nodes such as the collection point, cold storage warehouse, and processing workshop on the beef transportation path, the graph neural network is used to construct a topological relationship diagram. In the fruit transportation, between the collection center and the transfer station, if the load margin value is -0.27 (calculated above), assuming that the transportation time of this transportation section is half a day (0.5 days), then the weight of the edge between them is -0.27×0.5=-0.135. From the transfer station to the distribution center, if the load factor of the transfer station is 0.5 (assumed), the load factor of the distribution center is 0.4, the load margin value is 1-(0.5+0.4)=0.1, and the transportation time is 1 day, then the edge weight is 0.1×1=0.1. For beef transportation, from the farm collection point to the cold storage warehouse, if the bearing margin value is -0.292 and the transportation time is 1 day, the edge weight is -0.292×1=-0.292; from the cold storage warehouse to the processing workshop, assuming that the bearing margin value is 0.05 and the transportation time is 0.5 days, the edge weight is 0.05×0.5=0.025. The topological relationship diagram constructed in this way can reflect the closeness of the relationship between each node. The edge weight comprehensively considers the bearing margin value and transportation time, which not only reflects the remaining bearing capacity of the transportation section, but also takes into account the transportation time factor, providing a basis for finding the optimal transportation path.

[0085] Step S1413: Using a dynamic programming algorithm, search the topology diagram for an optimal path set that meets a temperature fluctuation tolerance threshold.

[0086] In the cold chain transportation of fruits, taking banana transportation as an example, the temperature fluctuation tolerance threshold of bananas is ±0.5℃ between 12-14℃. Starting from the collection center of the fruit production area, passing through multiple transfer stations and finally reaching the urban sales market, the dynamic programming algorithm is used to search for paths based on the constructed topological relationship diagram. Assume that there are two paths from the collection center to the first transfer station. One path passes through the mountainous area. Although the transportation time is short, the bumpy road may affect the operation of the refrigeration equipment and cause large temperature fluctuations; the other path passes through the plains. The transportation time is slightly longer but the temperature is easier to control. The dynamic programming algorithm will comprehensively consider the node load margin value, edge weight and temperature fluctuation on each path. If the path through the mountainous area has a short transportation time, but the temperature fluctuation may exceed the temperature fluctuation tolerance threshold of bananas, then this path may not be selected. The path through the plains, although the transportation time is slightly longer, can ensure that the temperature is within the tolerance threshold, and the node load margin value and edge weight also meet the requirements, then this path will be included in the optimal path set. For beef transportation, the temperature fluctuation tolerance threshold of beef is ±0.3℃ between -2℃ and 0℃. When searching for the optimal path from the farm to the processing enterprise in the topological relationship diagram, the conditions of each node and the temperature control during transportation should also be considered. If the refrigeration equipment of the cold storage warehouse on a certain path is unstable, which may cause the temperature fluctuation to exceed the tolerance threshold, then this path does not meet the requirements, and those paths that can ensure that the temperature is within the tolerance threshold and perform well in terms of node load margin value and edge weight will be selected into the optimal path set.

[0087] Step S1414, performing a Monte Carlo simulation on each optimal path in the optimal path set, and selecting an optimal path with a failure risk lower than a preset threshold as a trunk transportation channel.

[0088] In terms of fruit transportation, Monte Carlo simulation is performed for each path in the optimal path set for banana transportation. Assume that the preset threshold is 10%. An optimal path may pass through multiple nodes, such as a collection center, a transfer station, and a distribution center. In the Monte Carlo simulation, the historical failure rate of each node is considered. The refrigeration equipment of the collection center failed 3 times in the past year, with a total operating days of 300 days and a failure rate of 3 / 300=1%; the temperature fluctuation was abnormal twice due to power supply problems in the transfer station in the past year, with a total operating days of 300 days and a failure rate of 2 / 300≈0.67%; the distribution center had four transportation efficiency issues due to loading and unloading equipment failures in the past year, with a total operating days of 300 days and a failure rate of 4 / 300≈1.33%. Combining the failure rates of each node on this path, the total failure risk is calculated to be 1%+0.67%+1.33%=3%, which is lower than the preset threshold of 10%. This path can be used as a trunk transportation channel. Monte Carlo simulation is also performed for the optimal path for beef transportation. For example, for an optimal route from a farm collection point to a processing enterprise, the historical failure rate of the collection point is 2%, the historical failure rate of the intermediate cold storage warehouse is 1.5%, and the historical failure rate of the processing workshop is 0.5%. The total failure risk is 2%+1.5%+0.5%=4%, which is lower than the preset threshold of 10% and can also be used as a trunk transportation channel. Selecting the optimal route with low failure risk as the trunk transportation channel through Monte Carlo simulation can improve the reliability of cold chain transportation of fruits and beef.

[0089] And, step S142 includes: Step S1421, real-time collection of the in-cabin temperature data stream and external environment sensor data of the transport vehicle.

[0090] During the cold chain transportation of fruits and beef, for fruit transport vehicles, temperature sensors are installed in the vehicle to collect the temperature data stream in the cabin in real time. For example, during the transportation of bananas, the sensor records the temperature data at a certain interval (such as 5 minutes). At the same time, environmental sensors are also installed on the outside of the vehicle to collect external temperature, humidity, air pressure and other data. During the transportation of beef, the temperature sensor in the cabin of the transport vehicle accurately monitors the temperature of the environment in which the beef is located, and also records the data at a certain time interval. The external environmental sensor also obtains the external environmental information in real time. Taking banana transportation as an example, if it is transported in summer, the external environmental temperature may be as high as 30℃ or above, and the humidity is high, which forms a large temperature difference with the 12-14℃ temperature set in the cabin to keep the bananas fresh. For beef transportation, in the cold winter, the external environmental temperature may be as low as -10℃ or below, and the cabin temperature must be maintained at -2-0℃. These real-time collected data provide the basis for subsequent analysis and decision-making.

[0091] Step S1422, calculating the deviation between the actual temperature fluctuation value of the current transport section and the predicted tolerance threshold.

[0092] In fruit transportation, for banana transportation, if the predicted temperature fluctuation tolerance threshold is ±0.5℃ between 12-14℃, that is, 11.5-14.5℃. During transportation, the real-time collected cabin temperature data shows that the temperature reaches a minimum of 11℃ and a maximum of 15℃ in a certain period of time, then the actual temperature fluctuation range is 11-15℃. When calculating the deviation, first determine the range that exceeds the tolerance threshold, the lower limit exceeds 0.5℃ (11-11.5), and the upper limit exceeds 0.5℃ (15-14.5). The deviation can be calculated by a certain algorithm (such as the ratio of the sum of the temperature values ​​exceeding the range to the tolerance threshold range, etc.). For beef transportation, if the predicted temperature fluctuation tolerance threshold is ±0.3℃ between -2-0℃, that is, -2.3-0.3℃, the actual collected temperature data shows a minimum of -2.5℃ and a maximum of 0.5℃, then the lower limit exceeds 0.2℃ and the upper limit exceeds 0.2℃, and the deviation is also calculated according to the corresponding algorithm.

[0093] Step S1423: When the deviation exceeds the first critical value, the standby refrigeration equipment is activated and the transportation time compensation coefficient is recalculated.

[0094] In the transportation of bananas, assuming that the first critical value is set to 10% (the 10% here is a relative value obtained according to the deviation calculation method), if the previously calculated deviation exceeds 10%, for example, it reaches 15%, it means that the temperature fluctuation has had a significant impact on the preservation of bananas. At this time, the standby refrigeration equipment is activated. After the standby refrigeration equipment is started, the refrigeration power is increased to pull the temperature back to the tolerance threshold as soon as possible. At the same time, due to the startup of the refrigeration equipment and the temperature adjustment, the transportation time may be affected, and the transportation time compensation coefficient needs to be recalculated. The original transportation time critical value of bananas is 3 days. If the transportation time critical value may be shortened to 2.5 days due to temperature fluctuations and refrigeration equipment adjustments, it is necessary to recalculate factors such as speed and stop time during transportation to ensure that the bananas arrive at the destination within the specified new transportation time and the quality is not affected. For beef transportation, if the deviation exceeds the first critical value, for example, the first critical value of beef is set to 8%, and the actual calculated deviation is 10%, the standby refrigeration equipment is activated and the transportation time compensation coefficient is recalculated, such as adjusting the original 2-day transportation time critical value to 1.8 days to ensure the freshness of the beef.

[0095] Step S1424: When the deviation exceeds the second critical value, the path switching mechanism is triggered and the emergency storage allocation priority is updated.

[0096] In the transportation of bananas, assuming that the second critical value is set to 20%, if the calculated deviation exceeds 20%, for example, reaching 25%, this indicates that the temperature control problem on the current transportation path is very serious, which may be due to a refrigeration equipment failure or a harsh external environment with a continuous impact. At this time, the path switching mechanism is triggered to find an alternative path in the previously constructed cold chain transportation path topology. At the same time, the emergency storage allocation priority is updated. If there are multiple emergency storage points to choose from, the emergency storage point that is closer to the current location and has better temperature control conditions is given priority. The bananas are transferred to the emergency storage point for inspection and temperature adjustment to prevent further deterioration of the bananas. For beef transportation, if the deviation exceeds the set second critical value (such as 15%) and reaches 18%, the path switching mechanism is triggered to select other suitable transportation paths, and the emergency storage allocation priority is updated to transfer the beef to the emergency storage point for processing to ensure that the quality of the beef is not further affected.

[0097] In a possible implementation, step S143 includes: Step S1431, spatially matching the weather forecast data with the geographical locations of the nodes in the path topology.

[0098] In this embodiment, the transportation path topology of bananas and beef is taken as an example, which contains many nodes, such as the banana production collection center, the transfer station during transportation, the fruit market at the destination, the beef farm collection point, the cold storage warehouse, the processing enterprise, etc. The meteorological forecast data covers a variety of meteorological elements such as temperature, humidity, wind speed, and precipitation. For example, the meteorological department provides detailed meteorological forecast data for the next 24 hours, including that a certain area will experience rain and cooling weather. In the path topology, the collection center is located in a specific latitude and longitude coordinate area, and the transfer station is located in another area. The geographical location information in the meteorological forecast data is accurately matched with the geographical location of the node in the path topology. If the meteorological forecast in the area where the collection center is located shows that rainfall will cause water accumulation on the road, which may affect the efficiency of vehicle access and cargo loading and unloading, this meteorological information is associated with the node of the collection center. For beef transportation, if the area where the cold storage warehouse is located is predicted to have strong winds, it may affect the stability of the power supply of the cold storage warehouse, and this meteorological information is matched with the cold storage warehouse node. Through this spatial matching, the potential impact of different meteorological conditions on each node of the cold chain transportation path can be clearly defined.

[0099] Step S1432, identifying extreme weather areas that may affect the transportation route within a preset time period in the future.

[0100] Continuing with the example of fruit and beef transportation, analysis is performed based on weather forecast data within a preset time period in the future (such as the next 48 hours). For banana transportation, if weather data shows that a section of mountainous road on a certain transportation route is about to encounter heavy rain and thunderstorms, this mountainous area is identified as an extreme weather area that may affect the transportation route. Because heavy rain may cause landslides and muddy roads, and lightning may affect the operation of the vehicle's electronic equipment, thereby threatening the cold chain transportation safety of bananas. In terms of beef transportation, if a plain area is predicted to have heavy snow weather, this plain area is an extreme weather area. Heavy snow may cause snow and ice on the road, affecting the speed and safety of the transport vehicle, and may also affect the timeliness of beef transportation, because beef needs to arrive at the processing company within the specified time to ensure freshness. Through the analysis of meteorological data, these extreme weather areas can be accurately identified so that corresponding countermeasures can be taken.

[0101] Step S1433, prioritizing all in-transit transport tasks of the affected path segment, wherein the priority ranking is based on the cargo spoilage rate, the remaining transport time and the availability of alternative paths.

[0102] In the transportation of fruits, for the transportation of bananas, it is assumed that there are multiple in-transit transportation tasks on a certain transportation route. If one batch of bananas has been transported for two days and there is still one day to the destination, and the bananas have a relatively fast corruption rate, under normal circumstances, the remaining transportation time can just ensure the freshness of the bananas, but because the route segment is affected by extreme weather, the priority needs to be re-evaluated. At the same time, the availability of alternative routes must also be considered. If there is an alternative route, although it may be slightly longer, it can avoid the extreme weather area, then the priority of this task is relatively high. For another batch of transportation tasks that will arrive at the destination in two days and have a slower corruption rate of bananas, if there is no suitable alternative route, its priority is relatively low. In the transportation of beef, if a batch of beef has been transported most of the way, the remaining transportation time is short, the beef has a faster corruption rate, and there is an alternative transportation route, then the priority of this in-transit transportation task is higher. If a batch of beef has a long remaining transportation time, a slower corruption rate, and no alternative route, its priority is lower. By comprehensively considering factors such as the cargo corruption rate, the remaining transportation time, and the availability of alternative routes, all in-transit transportation tasks on the affected route segments are reasonably prioritized.

[0103] Step S1434, generating a time-divided transport resource reallocation instruction set, wherein the transport resource reallocation instruction set specifies the speed regulation strategy and refrigeration power adjustment range of each cold chain vehicle in each time window.

[0104] Take the cold chain transportation of fruits and beef as an example. In the process of banana transportation, a time-divided transport resource reallocation instruction set is generated based on the previous analysis results. If a certain transportation route segment is affected by extreme weather, such as rainy weather, during a certain period of time. For cold chain vehicles on this route segment, in a time window (such as 2-3 hours) before the start of rainfall, due to slippery roads, in order to ensure safety, the transport resource reallocation instruction set may specify that the vehicle reduce the driving speed, for example, from the original 60 kilometers per hour to 40 kilometers per hour. At the same time, due to the reduced speed, the heat exchange between the vehicle and the outside world is reduced, and the refrigeration power adjustment range of the refrigeration equipment may be reduced by 10% to avoid over-refrigeration. In beef transportation, when facing blizzard weather, in a time window (such as 1-2 hours) before the blizzard arrives, the transport resource reallocation instruction set may require the vehicle to reduce the speed in advance, from 80 kilometers per hour to 50 kilometers per hour, and due to the decrease in external temperature, in order to maintain the temperature in the vehicle at -2-0℃, the refrigeration power adjustment range may be increased by 15% to prevent the beef from being frozen. As time goes by, different time windows continuously adjust the speed control strategy and refrigeration power adjustment range of each cold chain vehicle according to meteorological conditions and the actual situation of the transportation task to ensure the quality and safety of fruits and beef during the cold chain transportation process.

[0105] In a possible implementation, the method further includes: Step S210: dividing the target area into a plurality of cold chain demand sub-areas based on the historical order distribution density.

[0106] In this embodiment, in the cold chain transportation scenario of fruits and beef, the target area is considered to be a large area covering multiple cities and their surrounding rural areas. The statistical data of historical order distribution density comes from the cold chain transportation order records of the past years. For example, for the transportation orders of bananas and beef, after detailed analysis, it is found that the order density is very high near the commercial district in the city center due to the presence of many supermarkets, restaurants and retailers. These areas have a large and stable demand for fruits and beef, and there is a large demand for cargo transportation every day. In some small residential areas on the outskirts of the city, the order density is relatively low, and the demand for cargo transportation is mainly concentrated on weekends or holidays because residents will make centralized purchases at these times. In rural areas, near fruit plantations and beef farms, the order density shows cyclical changes related to the harvest season. During the fruit harvest season, such as banana harvest, the number of transportation orders from the periphery of the plantation to the city increases significantly, while it is less in the non-harvest season.

[0107] Based on such historical order distribution density, the target area is divided into multiple cold chain demand sub-areas. The central business district of the city can be divided into a sub-area, marked as sub-area A. The cold chain demand here is characterized by stable and large transportation volume, and extremely high requirements for the freshness and timeliness of goods, because commercial activities in these areas rely on the continuous supply of fresh fruits and beef. Small residential areas in the suburbs of the city are divided into sub-area B, whose cold chain demand is cyclical and volatile, with relatively small transportation volume, and the requirements for the freshness of goods depend to a certain extent on the consumption habits of residents. Rural areas near plantations and farms are further subdivided according to different crops and breeding types, such as sub-area C around banana plantations and sub-area D around beef farms. Sub-area C has a strong demand for cold chain during the banana harvest season, and bananas need to be transported out quickly to avoid large accumulation and deterioration, while the demand drops sharply in the non-harvest season. The beef transportation demand in sub-area D is related to the breeding cycle and market demand. When the breeding is mature and the market demand is high, the demand for cold chain transportation is large.

[0108] Step S220, establishing an independent demand fluctuation baseline for each cold chain demand sub-area, wherein the demand fluctuation baseline includes daily minimum transportation volume, typical cargo composition ratio, and peak period characteristics.

[0109] For sub-area A, the minimum daily transportation volume is calculated based on historical data. For example, at least 100 boxes of bananas and 500 kilograms of beef need to be transported every day to meet basic commercial needs. In terms of the typical cargo composition ratio, since this is a commercial area, the sales ratio of fruit and beef is relatively balanced, with bananas accounting for 40% of the total cargo transportation volume and beef accounting for 60%. The peak time period is characterized by the lunch and dinner time on weekdays and all day on weekends. During these periods, the demand for fruit and beef in supermarkets, restaurants, etc. increases significantly, and the transportation demand reaches a peak.

[0110] In sub-area B, the minimum daily transport volume is relatively small, perhaps 20 boxes of bananas and 100 kilograms of beef per day. In the typical cargo composition ratio, due to residents' consumption habits, the demand for fruit may be slightly higher than beef, with bananas accounting for 60% and beef accounting for 40%. The peak hours are mainly concentrated in the afternoons and evenings on weekends, when residents have more time to shop and the demand for fruit and beef increases.

[0111] In sub-area C, during the banana harvest season, the minimum daily transportation volume may be as high as 500 boxes of bananas, and the typical cargo composition ratio is almost 100% bananas. The peak period is the concentrated transportation period after the banana harvest, which usually lasts for several weeks. In the non-harvest season, the minimum daily transportation volume may drop below 10 boxes, and the typical cargo composition ratio is still 100% bananas, but the transportation demand is extremely low. Around the beef farm in sub-area D, the minimum daily transportation volume depends on the scale of breeding and market orders. It is assumed to be 300 kilograms of beef per day, and the typical cargo composition ratio is 100% beef. The peak period is related to the breeding cycle, such as the concentrated slaughter period after the breeding matures and the peak market demand during holidays.

[0112] Step S230: When the deviation between the sub-region order flow monitored in real time and the demand fluctuation baseline exceeds a preset threshold, the inter-region capacity balancing mechanism is triggered.

[0113] In actual cold chain transportation operations, the order flow of each sub-region is continuously tracked through a real-time monitoring system. Taking sub-region A as an example, if on a certain day, the order flow of bananas is monitored to be only 50 boxes, which is far below the demand fluctuation baseline of the daily minimum transportation volume of 100 boxes, and the order flow of beef is 300 kilograms, which is also below the demand fluctuation baseline of 500 kilograms. Assuming that the preset threshold is 30%, the deviation of the banana order flow is (100-50) / 100=50%, and the deviation of the beef order flow is (500-300) / 500=40%, both exceeding the preset threshold.

[0114] For sub-region C during the banana harvest season, if for some reason, such as a transport vehicle failure or bad weather, the daily transport volume can only reach 300 boxes, while the demand fluctuation baseline is 500 boxes, the deviation is (500-300) / 500=40%, exceeding the preset threshold. When this happens, the inter-regional transport capacity balancing mechanism is triggered.

[0115] Among them, the capacity balancing mechanism includes cold storage resource sharing agreements in adjacent sub-regions, dynamic scheduling rules for cross-regional transport vehicles, and temporary storage facility activation strategies.

[0116] When the order flow in sub-region A is lower than the demand fluctuation baseline, according to the cold storage resource sharing agreement of the neighboring sub-regions, sub-region A can share cold storage resources with the neighboring sub-region B. For example, the cold storage in sub-region B has a certain amount of idle capacity during non-peak hours, and sub-region A can temporarily store some of the goods originally planned to be transported to its own cold storage in the cold storage of sub-region B to alleviate its own transportation pressure.

[0117] In terms of dynamic dispatching rules for inter-regional transport vehicles, if sub-region C has insufficient transport capacity during the banana harvest season, and transport vehicles in sub-region A have surplus transport capacity after meeting their own needs, some vehicles in sub-region A can be dispatched to sub-region C for banana transportation according to the dynamic dispatching rules for inter-regional transport vehicles. These vehicles need to operate according to the specified time, temperature requirements and transport routes to ensure the cold chain transportation quality of bananas.

[0118] The strategy of activating temporary storage facilities may also work in sub-area D. If sub-area D encounters a shortage of transportation capacity during the beef delivery period, and there are temporary storage facilities nearby, these facilities can be activated. Some beef can be temporarily stored in temporary storage facilities, waiting for the dispatch of transportation vehicles, to prevent the beef from spoiling while waiting for transportation.

[0119] In a possible implementation, the method further includes: Step S310, calculating a transportation time efficiency association matrix between each demand fluctuation sub-region, wherein the matrix elements of the transportation time efficiency association matrix represent the maximum allowable transportation time between two demand fluctuation sub-regions that meets the temperature requirements.

[0120] Taking sub-regions A, B, C, and D as examples, calculate the transportation time correlation matrix between them. From sub-region A to sub-region B, due to the short distance and convenient transportation, the road conditions are good, and the temperature requirements of fruits and beef can be guaranteed during transportation, the maximum allowable transportation time may be 2 hours. From sub-region A to sub-region C, the distance is relatively far, and different climate zones may be passed on the way, requiring more complex temperature control measures, and the maximum allowable transportation time is 4 hours. From sub-region A to sub-region D, considering the temperature sensitivity of beef and the conditions of the transportation route, the maximum allowable transportation time is 3 hours.

[0121] From sub-area B to sub-area C, the maximum allowable transportation time is 3 hours due to the need to cross urban and rural areas and the complex traffic conditions. From sub-area B to sub-area D, the maximum allowable transportation time is 2.5 hours. From sub-area C to sub-area D, although the distance is not far, the different temperature requirements of bananas and beef and the mutual influence during transportation must be considered, so the maximum allowable transportation time is 2 hours. These matrix elements constitute the transportation time association matrix, which provides an important basis for the planning and scheduling of cold chain transportation.

[0122] Step S320: updating the time parameters in the transport time dependency matrix according to the real-time traffic flow data.

[0123] During the actual transportation process, real-time traffic flow data is constantly changing. For example, during the morning and evening rush hours on weekdays, the roads inside the city are congested. The traffic flow from sub-region A to sub-region B increases, and the road speed decreases. The originally maximum allowed transportation time is 2 hours, but due to traffic congestion, this time may need to be extended to 3 hours. At night, the traffic flow decreases and the roads are clear. The transportation time from sub-region A to sub-region C may be shortened from 4 hours to 3.5 hours.

[0124] From sub-region B to sub-region D, if there is a traffic accident or road construction, the traffic flow will be severely affected. The maximum allowed transportation time may increase from 2.5 hours to 4 hours. By monitoring and analyzing real-time traffic flow data, the time parameters in the transportation timeliness correlation matrix are updated in a timely manner to ensure that cold chain transportation can be reasonably arranged according to the actual situation.

[0125] Step S330, when the actual transportation time between any two demand-fluctuating sub-regions exceeds the corresponding value in the transportation timeliness correlation matrix, generate a detour route suggestion and evaluate the additional energy consumption cost.

[0126] Suppose during the transportation process, the actual transportation time from sub-region A to sub-region C exceeds 4 hours in the transportation timeliness correlation matrix. At this time, the system will generate a detour route suggestion. The original route may pass through the congested section in the city center, and now it is recommended to detour through the roads on the outer ring of the city. However, the detour route may be longer and require more fuel consumption and running time of the refrigeration equipment.

[0127] For the detour route, it is necessary to evaluate the additional energy consumption cost. If the energy consumption cost of the original route is 1 yuan per kilometer (including the energy consumption of the refrigeration equipment), the total distance is 100 kilometers, and the total energy consumption cost is 100 yuan. The distance of the detour route becomes 120 kilometers. Due to the changes in road conditions and driving speed, the energy consumption cost per kilometer may become 1.2 yuan, and the total energy consumption cost becomes 144 yuan. The additional energy consumption cost is 144 - 100 = 44 yuan.

[0128] Step S340, compare the detour route suggestion with the economic priority of the current transportation task through weighting, and select the path adjustment plan with the best cost-benefit.

[0129] For the transportation task from sub-region A to sub-region C, if it is transporting high-value beef and the current economic priority is high, emphasizing the freshness and timely arrival of the goods. Although the detour route has additional energy consumption costs, it can ensure that the beef reaches the destination within the specified temperature and time, avoiding greater economic losses caused by spoilage. Then, even if the cost of the detour route increases, from the perspective of cost-benefit, the detour route may still be selected.

[0130] If the bananas transported are relatively low in value and have a low economic priority, the requirements for transportation time are not particularly strict. After comparing the additional energy consumption cost of the detour route and the possible losses caused by the slight deterioration of the bananas, if the additional energy consumption cost is too high, it may be chosen to continue transporting along the original route, while taking some measures, such as adjusting the power of the refrigeration equipment, to minimize the risk of banana deterioration. Through such a weighted comparison, the most cost-effective route adjustment plan that best suits the current transportation task is selected, and the economic optimization of cold chain transportation is achieved while ensuring the quality of the goods.

[0131] In a possible implementation, the method further includes: Step S410, establishing a full life cycle performance attenuation model for cold chain transportation equipment, wherein the full life cycle performance attenuation model includes a compressor working efficiency curve, an insulation material aging coefficient, and a battery capacity attenuation function.

[0132] In this embodiment, in the cold chain transportation scenario of fruits and beef, cold chain transportation equipment is essential to ensure the quality of goods. Taking the transport vehicle as an example, the refrigeration equipment in the vehicle includes a compressor, and its working efficiency changes continuously throughout the life cycle. When the equipment is newly put into use, the compressor has a high working efficiency and can quickly reduce the temperature in the car to the temperature range required for fruits (such as 12-14°C for bananas) and beef (-2-0°C). As the use time increases and the number of transportations increases, the working efficiency of the compressor gradually decreases. For example, during the first 1000 hours of operation, the compressor can reduce the temperature of the car from room temperature to the target temperature at a rate of 5°C per hour. But when the operating time reaches 5000 hours, this cooling rate may be reduced to 3°C per hour. Through long-term data collection and analysis, a compressor working efficiency curve can be drawn, which reflects the changes in the working efficiency of the compressor under different operating times.

[0133] The aging coefficient of insulation materials is also an important part of the performance attenuation model of the entire life cycle. The compartments of cold chain transport vehicles use insulation materials to reduce heat exchange. Over time, the insulation materials will gradually age. New insulation materials can effectively prevent external heat from entering the compartment during transportation, so that the energy consumption of refrigeration equipment is kept at a relatively low level. For example, in the early stage of use, bananas with a temperature of 13°C in the compartment are transported for 1 hour in an environment with an external temperature of 30°C, and the temperature rise does not exceed 0.5°C. But after several years of use, the insulation material ages, and the temperature rise may reach 1°C under the same conditions. According to the changes in the insulation performance of the insulation material under different usage times, the aging coefficient of the insulation material can be determined.

[0134] For electric cold chain transport vehicles, the battery capacity decay function is a factor that must be considered. New batteries can provide sufficient power support for refrigeration equipment and vehicle driving. As the number of charging and discharging times increases, the battery capacity gradually decreases. For example, a new battery can support the vehicle to travel continuously for 500 kilometers and maintain the normal operation of the refrigeration equipment after being fully charged. After 500 charge and discharge cycles, the battery's range may be reduced to 400 kilometers, and the operating time of the refrigeration equipment will be shortened accordingly. By monitoring and analyzing the capacity changes of the battery under different charge and discharge times, a battery capacity decay function is established. These compressor working efficiency curves, insulation material aging coefficients, and battery capacity decay functions together constitute a full life cycle performance decay model for cold chain transport equipment, providing an important basis for equipment maintenance and management.

[0135] Step S420: predicting the maintainable time window of each transport vehicle based on the real-time equipment monitoring data.

[0136] During the cold chain transportation process, various sensors are installed on the vehicles to monitor the operating status of the equipment in real time. For refrigeration equipment, sensors can monitor the operating pressure, temperature, current and other parameters of the compressor. Taking the operating pressure of the compressor as an example, under normal circumstances, the operating pressure of the compressor should be maintained within a stable range, such as 2-3 MPa. When the real-time monitoring shows that the operating pressure of the compressor gradually increases and exceeds the normal range, and this trend continues for a period of time, this may be a signal of wear of the internal components of the compressor or blockage of the refrigeration system. Combined with the compressor working efficiency curve in the full life cycle performance attenuation model, the analysis found that when the operating pressure reaches this abnormal value, according to historical data and model predictions, the compressor may still work normally for 100 hours, but serious failures may occur afterwards.

[0137] For thermal insulation materials, sensors can monitor the temperature difference between the inside and outside of the car. If the temperature difference between the inside and outside of the car is found to gradually decrease under the same external environment, this indicates that the thermal insulation performance of the thermal insulation material is decreasing. According to the thermal insulation material aging coefficient model, when the temperature difference decreases to a certain extent, it is predicted that the thermal insulation material may be able to maintain the current state for 300 hours, after which it will need maintenance or replacement.

[0138] For the battery of an electric vehicle, the battery voltage, current, and remaining power are monitored. When the battery voltage drops rapidly after charging, or the remaining power is consumed much faster than normal, combined with the battery capacity decay function, it is predicted that the battery may still support the normal operation of the vehicle for 200 hours. This is the battery's maintainable time window. By analyzing these real-time equipment monitoring data and combining the full life cycle performance decay model, the maintainable time window of different equipment on each transport vehicle can be accurately predicted.

[0139] Step S430, embedding a preventive maintenance plan in the dynamic scheduling plan, wherein the preventive maintenance plan specifies a priority maintenance period for each vehicle and an alternative vehicle scheduling plan.

[0140] In the cold chain transportation of fruits and beef, a preventive maintenance plan is formulated based on the forecast of the maintainable time window of each transport vehicle. For a vehicle transporting bananas, it is predicted that the compressor of its refrigeration equipment has a maintainable time window of 100 hours, the insulation material has 300 hours, and the battery has 200 hours. Considering the importance of the compressor in maintaining the appropriate temperature of bananas, the priority maintenance period can be set to a low transportation demand period within the next 100 hours. For example, in the off-season of banana transportation, the transportation demand is low, and the vehicle can be maintained during this period.

[0141] During the maintenance period, a replacement vehicle dispatch plan needs to be developed. If the vehicle was originally responsible for the transportation task from the banana plantation to the city wholesale market, a spare vehicle will be deployed from other existing transport vehicles. This spare vehicle needs to meet the same cold chain transportation requirements as the original vehicle, including refrigeration capacity, compartment capacity, etc. During the dispatch process, the transportation route and schedule must be adjusted to ensure that the bananas can reach the destination within the specified temperature and time. The same is true for vehicles transporting beef. The priority maintenance period is determined according to the maintainable time window of different equipment, and a suitable replacement vehicle is arranged to ensure uninterrupted cold chain transportation of beef.

[0142] Step S440: When the predicted value of the equipment failure probability exceeds the safety threshold, the maintenance plan is executed in advance and the transportation path topology is recalculated.

[0143] In the cold chain transportation process, let's continue to use the example of vehicles transporting bananas and beef, assuming that the safety threshold is set at a 10% probability of equipment failure. Through real-time analysis of equipment operation data and prediction of the full life cycle performance degradation model, if the predicted value of the compressor failure probability of a vehicle's refrigeration equipment reaches 15%, it exceeds the safety threshold. At this time, regardless of whether it is currently in the priority maintenance period, the maintenance plan must be executed in advance.

[0144] Since the vehicle needs maintenance, the transportation route topology originally planned based on this vehicle needs to be recalculated. For example, this vehicle was originally a key link in the transportation route from the fruit production area through multiple transfer stations to the urban sales market. When recalculating the transportation route topology, it is necessary to consider other available transportation vehicles, the load factor of each node, the transportation time, and the temperature fluctuation tolerance threshold and other multi-dimensional cold chain demand parameters. The transportation route may be adjusted, other vehicles may be selected, or transportation tasks for certain nodes may be increased to ensure that the cold chain transportation of bananas and beef can continue to be carried out efficiently and safely.

[0145] In a possible implementation, the method further includes: Step S510, constructing a multi-level cold chain resource allocation strategy, wherein the multi-level cold chain resource allocation strategy includes a core node resource reservation mechanism and an edge node dynamic sharing mechanism.

[0146] In the cold chain transportation network of fruits and beef, there are multiple levels of nodes. Core nodes can be large fruit collection centers, storage centers of beef processing companies, etc. These nodes play a key hub role in cold chain transportation. Edge nodes are some small transfer stations, temporary storage points at the retail end, etc.

[0147] The core node resource reservation mechanism is to ensure that the core nodes can meet important transportation needs at critical moments, such as the fruit harvest season or the peak beef sales season. Take the fruit collection center as an example. During the banana harvest season, the collection center needs to reserve a certain proportion of storage space, loading and unloading equipment, and transportation vehicle resources. Assuming that the total storage space of the collection center is 1,000 cubic meters, based on historical data and predicted harvest volume, 30% of the storage space, or 300 cubic meters, is reserved specifically for the possible arrival of a large number of bananas. For loading and unloading equipment, 20% of the equipment working time is reserved to ensure efficient loading and unloading of bananas during peak hours. For transportation vehicles, 10% of vehicle capacity is reserved so that bananas can be transported out in time in an emergency.

[0148] The edge node dynamic sharing mechanism takes into account the flexibility of edge node resources. For example, at a small transfer station, while its storage space and refrigeration equipment meet its own daily transportation tasks, if the adjacent retail temporary storage point has a backlog of goods and needs more storage space, according to the edge node dynamic sharing mechanism, part of the transfer station's idle storage space can be shared with the retail temporary storage point. Similarly, for refrigeration equipment, if the retail refrigeration equipment fails, the transfer station can provide refrigeration support to a certain extent to ensure that the quality of fruit and beef is not affected.

[0149] Step S520, calculating the resource gap of nodes at each level according to the real-time demand forecast result.

[0150] In the cold chain transportation of fruits and beef, the resource requirements of nodes at all levels are predicted in real time through the analysis of market demand, output, transportation plan and other data. Taking beef transportation as an example, the storage center of a beef processing enterprise is a core node. According to historical order data and market trend forecasts, it is expected that 1,000 kg of beef will need to be stored and transported every day during the upcoming holidays. However, the current maximum storage capacity of the storage center is 800 kg, and the transportation capacity of the transport vehicles can only meet the transportation of 800 kg per day. Therefore, the resource gap of the core node is 200 kg in storage and 200 kg in transportation.

[0151] For edge nodes, such as small transfer stations, it is predicted that 50 boxes of fruit need to be transferred every day in a certain period of time, but the actual transfer capacity is only 40 boxes, and the resource gap is 10 boxes. In this way, the resource gap of nodes at all levels can be accurately calculated to provide a basis for resource allocation.

[0152] Step S530: When the resource gap of the core nodes in the nodes at all levels exceeds the first threshold, the cross-regional cold storage resource allocation process is started.

[0153] Assume that the first threshold is set to 20% of the total resources of the core node. In fruit transportation, a large fruit collection center is a core node. During the peak fruit season, its storage resource gap reaches 30%, exceeding the first threshold. At this time, the cross-regional cold storage resource allocation process is started. If there are other cold storages with free storage space in the surrounding area, the excess fruit in this area can be allocated to these cold storages for storage. For example, due to insufficient storage space for bananas in the local collection center, some bananas can be transported to a cold storage in an adjacent area, which is 100 kilometers away from the local collection center. During transportation, it is necessary to ensure that the cold chain transportation conditions meet the temperature requirements of the bananas, and to coordinate the transportation plan and cost to avoid adverse effects on the quality and transportation efficiency of the bananas.

[0154] Step S540: When the resource gap of the edge nodes in the nodes at all levels exceeds the second threshold, a mobile temporary cold chain equipment deployment plan is triggered.

[0155] Assume that the second threshold is set to 30% of the total resources of the edge node. In beef transportation, a small transfer station is an edge node. Due to a sudden increase in transportation demand during a certain period, the capacity gap of its transport vehicles reached 40%, exceeding the second threshold. At this time, the mobile temporary cold chain equipment deployment plan is triggered. Mobile refrigerated trucks can be deployed to this transfer station to increase the transportation capacity of the transfer station. These mobile refrigerated trucks have independent refrigeration systems and a certain amount of storage space, which can supplement the insufficient capacity of the transfer station in a short time, ensuring that the beef can be transported to the next node in a timely and safe manner.

[0156] Step S550, feeding back the resource allocation result to the spatiotemporal prediction model in real time, and constructing a closed-loop optimization system for demand prediction and resource allocation.

[0157] During the cold chain transportation of fruits and beef, after the cross-regional cold storage resource allocation or the deployment of mobile temporary cold chain equipment is completed, the resource allocation results are fed back to the spatiotemporal prediction model in real time. For example, the data such as how many bananas are allocated from the local collection center to the cold storage in the adjacent area, the working hours of the mobile refrigerated trucks at the transfer station, and the transportation volume are fed back to the spatiotemporal prediction model. The spatiotemporal prediction model adjusts the future demand forecast based on these feedback data. If it is found that the cold chain transportation pressure in a certain area is relieved after the allocation of resources, and there are new changes in the transportation demand in another area, the resource allocation strategy can be readjusted. In this way, a closed-loop optimization system for demand forecasting and resource allocation is constructed to continuously improve the efficiency and reliability of cold chain transportation of fruits and beef.

[0158] Figure 2 A cold chain logistics service system 100 provided in an embodiment of the present application is shown, including a processor 1001 and a memory 1003 and a program code stored in the memory 1003, and the processor 1001 executes the above program code to implement the steps of the cold chain transportation demand analysis method based on big data prediction.

[0159] Figure 2 A cold chain logistics service system 100 provided in an embodiment of the present application is shown, including a processor 1001 and a memory 1003 and a program code stored on the memory 1003, and the processor 1001 executes the above program code to implement the steps of the cold chain transportation path optimization method based on AI prediction.

[0160] Figure 2 The cold chain logistics service system 100 shown includes: a processor 1001 and a memory 1003. The processor 1001 and the memory 1003 are connected, such as through a bus 1002. Optionally, the cold chain logistics service system 100 may also include a transceiver 1004, which can be used for data interaction between the cold chain logistics service system and other cold chain logistics service systems, such as data sending and / or data receiving. It should be noted that the transceiver 1004 is not limited to one in actual scheduling, and the structure of the cold chain logistics service system 100 does not constitute a limitation on the embodiments of the present application.

[0161] Processor 1001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. Processor 1001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0162] The bus 1002 may include a path to transmit information between the above components. The bus 1002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 1002 may be divided into an address bus, a data bus, a control bus, and the like.

[0163] The memory 1003 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), magnetic disk storage medium, other magnetic storage devices, or any other medium that can be used to have or store program code and can be read by a computer, without limitation herein.

[0164] The memory 1003 is used to store program codes for executing the embodiments of the present application, and the execution is controlled by the processor 1001. The processor 1001 is used to execute the program codes stored in the memory 1003 to implement the steps shown in the above method embodiments.

[0165] An embodiment of the present application provides a computer-readable storage medium having program code stored thereon. When the program code is executed by a processor, the steps and corresponding contents of the aforementioned method embodiment can be implemented.

[0166] It should be understood that, although each operation step is indicated by arrows in the flowchart of the embodiment of the present application, the implementation order of these steps is not limited to the order indicated by the arrows. Unless there is clear explanation in this article, in some implementation scenarios of the embodiment of the present application, the implementation steps in each flowchart can be performed in other orders based on demand. In addition, some or all of the steps in each flowchart may include multiple sub-steps or multiple stages according to the actual implementation scenario, and some or all of these sub-steps or stages may be executed at the same time, and each sub-step or stage in these sub-steps or stages may also be executed at different times respectively. In different scenarios at the execution time, the execution order of these sub-steps or stages can be flexibly configured based on demand, and the embodiment of the present application does not limit this.

[0167] The above is only an optional implementation method for some implementation scenarios of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the technical concept of the scheme of the present application, other similar implementation methods based on the technical ideas of the present application are also within the protection scope of the embodiments of the present application.

Claims

1. A cold chain transportation demand analysis method based on big data prediction, characterized in that: The method comprises: Acquire a historical cold chain order dataset of a target area, wherein the historical cold chain order dataset includes temperature-sensitive cargo transportation records and their corresponding environmental parameters over multiple time periods; Extracting a set of transportation demand features from the historical cold chain order data, wherein the set of transportation demand features includes a correlation feature between the type of goods and temperature sensitivity, a balance feature between transportation timeliness and equipment energy consumption, and a matching feature between regional climate and storage conditions; Input the transport demand feature set into a pre-trained spatiotemporal prediction model to generate multi-dimensional cold chain demand parameters for the target area within a future specified period, wherein the multi-dimensional cold chain demand parameters include a temperature fluctuation tolerance threshold, a transport timeliness critical value, and a path node load coefficient; Based on the multi-dimensional cold chain demand parameters, a dynamic scheduling plan for cold chain transportation with a timestamp is generated.

2. The cold chain transportation demand analysis method based on big data prediction according to claim 1 is characterized in that: The step of generating a dynamic cold chain transportation scheduling scheme with a timestamp based on the multi-dimensional cold chain demand parameters includes: Based on the multi-dimensional cold chain demand parameter optimization, a cold chain transportation path topology is generated, and each transportation node in the cold chain transportation path topology is associated with a dynamic resource allocation weight, and the dynamic resource allocation weight is negatively correlated with the historical failure rate of the corresponding node, the real-time temperature control capability, and the connectivity efficiency of adjacent nodes; According to the real-time operation status data of each node in the cold chain transportation path topology, the matching relationship between the transportation timeliness critical value and the temperature fluctuation tolerance threshold is dynamically adjusted to generate an abnormal event response strategy set, wherein the abnormal event response strategy set includes the activation conditions of the standby refrigeration equipment, the path switching priority rules and the emergency storage capacity allocation plan; The abnormal event response strategy set is spatially and temporally aligned with the real-time meteorological data stream to generate a dynamic scheduling plan for cold chain transportation with a timestamp. The dynamic scheduling plan for cold chain transportation includes an equipment pre-start time window, a transportation batch interval threshold, and a cross-regional collaborative transportation trigger condition.

3. The cold chain transportation demand analysis method based on big data prediction according to claim 1 is characterized in that: The step of extracting the transport demand feature set from the historical cold chain order data comprises: Identify the correspondence between the cargo category code in the historical cold chain order data and the temperature monitoring curve, and generate a temperature sensitivity distribution map; Analyze the correlation pattern between transportation time records and refrigeration equipment energy consumption logs, and establish a time efficiency-energy consumption balance coefficient matrix; Detect the matching degree between regional climate data and storage temperature maintenance capacity, and calculate the climate adaptability index; The temperature sensitivity distribution map, the time efficiency-energy consumption balance coefficient matrix and the climate adaptability index are integrated to generate a three-dimensional feature vector space as the transportation demand feature set; The training step of the spatiotemporal prediction model includes: Dividing the three-dimensional feature vector space into a training set and a validation set, and marking the training set and the validation set with corresponding time series labels; Constructing a deep spatiotemporal convolutional network, wherein the deep spatiotemporal convolutional network includes a climate fluctuation perception layer, a transportation path dependency layer, and an equipment state memory unit; Optimizing the network parameters of the deep spatiotemporal convolutional network by a back-propagation algorithm to minimize the spatiotemporal correlation loss function between the predicted output of the deep spatiotemporal convolutional network and the actual required parameters of the validation set; When the loss value change rate of N consecutive training cycles is lower than a preset threshold, the network parameters of the deep spatiotemporal convolutional network are frozen and the model weights are derived.

4. The cold chain transportation demand analysis method based on big data prediction according to claim 2 is characterized in that: The step of optimizing and generating a cold chain transportation path topology based on the multi-dimensional cold chain demand parameters comprises: Calculating the load margin value of each transport section according to the path node load factor; A topological relationship graph between transport nodes is constructed based on a graph neural network, wherein the edge weight in the topological relationship graph is determined by the product of the carrying margin value and the transport time efficiency; Using a dynamic programming algorithm to search for an optimal path set that meets a temperature fluctuation tolerance threshold in the topological relationship graph; Performing Monte Carlo simulation on each optimal path in the optimal path set, and selecting an optimal path with a failure risk lower than a preset threshold as a trunk transportation channel; And, the step of dynamically adjusting the matching relationship between the transportation timeliness critical value and the temperature fluctuation tolerance threshold value according to the real-time operation status data of each node in the cold chain transportation path topology to generate a set of abnormal event response strategies includes: Real-time collection of cabin temperature data streams and external environmental sensor data of transport vehicles; Calculate the deviation between the actual temperature fluctuation value of the current transport section and the predicted tolerance threshold; When the deviation exceeds the first critical value, the standby refrigeration equipment is activated and the transportation time compensation coefficient is recalculated; When the deviation exceeds the second critical value, the path switching mechanism is triggered and the emergency storage allocation priority is updated.

5. The cold chain transportation demand analysis method based on big data prediction according to claim 2 is characterized in that: The step of performing spatiotemporal alignment processing on the abnormal event response strategy set and the real-time meteorological data stream to generate a dynamic cold chain transportation scheduling plan with a timestamp includes: spatially matching the weather forecast data with the geographical locations of nodes in the path topology; Identify areas of extreme weather that may affect transportation routes within a preset time period in the future; Prioritize all in-transit transport tasks for the affected route segment based on factors including cargo spoilage rate, remaining transport time, and availability of alternative routes; Generate a time-divided transport resource reallocation instruction set, wherein the transport resource reallocation instruction set specifies the speed regulation strategy and refrigeration power adjustment range of each cold chain vehicle in each time window.

6. The cold chain transportation demand analysis method based on big data prediction according to claim 2 is characterized in that: The method further comprises: Divide the target area into multiple cold chain demand sub-areas based on historical order distribution density; Establish an independent demand fluctuation baseline for each cold chain demand sub-area, which includes the daily minimum transportation volume, typical cargo composition ratio and peak period characteristics; When the deviation between the sub-region order flow monitored in real time and the demand fluctuation baseline exceeds a preset threshold, the inter-regional capacity balancing mechanism is triggered; The transport capacity balancing mechanism includes a cold storage resource sharing agreement among neighboring sub-regions, dynamic scheduling rules for cross-regional transport vehicles, and a temporary storage facility activation strategy.

7. The cold chain transportation demand analysis method based on big data prediction according to claim 6 is characterized in that: The method further comprises: Calculate the transportation time efficiency association matrix between each demand fluctuation sub-region, wherein the matrix elements of the transportation time efficiency association matrix represent the maximum allowable transportation time between two demand fluctuation sub-regions that meets the temperature requirements; Update the time parameters in the transport time association matrix according to the real-time traffic flow data; When the actual transportation time between any two demand fluctuation sub-areas exceeds the corresponding value in the transportation time-efficiency correlation matrix, a detour route suggestion is generated and the additional energy consumption cost is evaluated; The detour route suggestions are compared with the economic priority of the current transportation task in a weighted manner, and the route adjustment plan with the best cost-effectiveness is selected.

8. The cold chain transportation demand analysis method based on big data prediction according to claim 2 is characterized in that: The method further comprises: Establish a full life cycle performance attenuation model for cold chain transportation equipment, wherein the full life cycle performance attenuation model includes a compressor working efficiency curve, an insulation material aging coefficient, and a battery capacity attenuation function; Predict the maintainable time window of each transport vehicle based on real-time equipment monitoring data; Embedding a preventive maintenance plan in the dynamic scheduling plan, the preventive maintenance plan specifying a priority maintenance period for each vehicle and an alternative vehicle scheduling plan; When the predicted value of equipment failure probability exceeds the safety threshold, maintenance planning is performed in advance and the transportation path topology is recalculated.

9. The cold chain transportation demand analysis method based on big data prediction according to claim 2 is characterized in that: The method further comprises: Constructing a multi-level cold chain resource allocation strategy, wherein the multi-level cold chain resource allocation strategy includes a core node resource reservation mechanism and an edge node dynamic sharing mechanism; Calculate the resource gap of nodes at all levels based on the real-time demand forecast results; When the resource gap of the core nodes in each level of nodes exceeds the first threshold, the cross-regional cold storage resource allocation process is started; When the resource gap of the edge nodes in the nodes at all levels exceeds the second threshold, the mobile temporary cold chain equipment deployment plan is triggered; The resource allocation results are fed back to the spatiotemporal prediction model in real time to build a closed-loop optimization system for demand prediction and resource allocation.

10. A cold chain logistics service system, characterized in that: It includes a processor and a computer-readable storage medium, wherein the computer-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are executed by the processor, the cold chain transportation demand analysis method based on big data prediction described in any one of claims 1 to 9 is implemented.

Images