Cross risk control method based on cross-regional logistics data

Abstract

The invention discloses a cross risk control method based on cross-regional logistics data, belongs to the technical field of logistics transportation, and realizes comprehensive representation of a complex logistics network by constructing a transportation directed graph of a multi-source heterogeneous fusion database. Secondly, multi-dimensional attribute values (time, economy, risk and replenishment) are introduced to comprehensively evaluate the connecting lines, so that the accuracy and adaptability of path selection are greatly improved; by setting the importance of different attributes and calculating the weight of the connecting line, the requirements of different logistics scenes can be flexibly met. Particularly, through calculation of estimation fitness and a dynamic selection mechanism based on real-time road conditions, changes of the logistics environment can be responded in time, the optimal connection line can be selected, the accuracy of path selection is improved, and the real-time optimization capability of the logistics process is enhanced. And finally, performing risk assessment on each connecting line experienced in the simulation process, and providing comprehensive risk insight for logistics management.

Description

Technical Field

[0001] The present invention relates to the field of logistics and transportation technology, and in particular to a cross-regional logistics data-based cross-risk control method. Background Art

[0002] Cold chain logistics is a vital part of modern supply chain management, involving the transportation and storage of temperature-sensitive products. Due to the strict requirements of cold chain products such as food and medicine on temperature, humidity and transportation time, mistakes in any link may lead to product quality degradation or loss. Therefore, risk management of cold chain logistics is particularly important. In cross-regional transportation, differences in geographical environment, climate change and infrastructure make risk factors more complex and diverse. Existing logistics risk control methods will comprehensively consider various indicators to obtain a highly comprehensive path, without considering demand conditions, and there is a problem of low path selection accuracy. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a cross-risk control method based on cross-regional logistics data to solve the problem of low path selection accuracy in the prior art.

[0004] In order to achieve the above-mentioned purpose of the invention, the technical solution adopted by the present invention is: a cross-risk control method based on cross-regional logistics data, comprising the following steps: S1. Construct a transportation directed graph based on a multi-source heterogeneous fusion database; S2. Generate a multidimensional attribute value for each connection line in the transport directed graph, wherein the multidimensional attribute value includes: a time attribute value, an economic attribute value, a risk attribute value, and a supply attribute value; S3. Set the importance of time, economy, risk and supply, and calculate the weight of the connection line based on the multi-dimensional attribute value; S4. Calculate the estimated fitness of the connection line according to the weight of the connection line; S5. On the transport directed graph, it is simulated that starting from the node to be shipped, based on the estimated fitness of the connection line and the real-time road conditions, at the intersection of each connection line, a connection line selection objective function is established to select the optimal connection line until the node to be received is reached, and the optimal route is obtained; S6. Evaluate logistics risks based on the connecting lines on the optimal route.

[0005] Furthermore, S1 is specifically as follows: taking each receiving location in the multi-source heterogeneous fusion database as a receiving node, taking each shipping location as a shipping node, mapping each path between the receiving location and the receiving node into a connecting line, and obtaining a transport directed graph.

[0006] Furthermore, the time attribute value in S2 is the normalized value of the truck passing time of each connecting line; The economic attribute value is a normalized value of the economic cost consumed on each connection line; The risk attribute value is a normalized value of the weighted values ​​of cargo damage risk, theft risk, traffic accident risk and natural disaster risk on each connection line; The supply attribute value is the normalized value of the number of supply stations on each connecting line.

[0007] Furthermore, S3 includes the following sub-steps: S31. Set the importance of time, economy, risk and supply, with the importance ranging from 0 to 1; S32. Construct a weight calculation function based on the importance and multi-dimensional attribute values ​​to obtain the weight of the connecting line.

[0008] Furthermore, the weight calculation function in S32 is: , where v is the weight of the connection line, t is the time attribute value of the connection line, E is the economic attribute value of the connection line, r is the risk attribute value of the connection line, s is the supply attribute value of the connection line, θ t is the importance of the time attribute value of the connection line, θ E is the importance of the economic attribute value of the connecting line, θ r is the importance of the risk attribute value of the connecting line, θ s is the importance of the replenishment attribute value of the connecting line.

[0009] Furthermore, the formula for calculating the estimated fitness of the connecting line in S4 is: , where h e is the estimated fitness of the connection line, α is the proportional coefficient of the weight of the connection line, v is the weight of the connection line, d is the straight-line distance from the node to be received after passing through the connection line, β is the proportional coefficient of the straight-line distance, and D is the maximum straight-line distance.

[0010] Furthermore, the S5 comprises the following sub-steps: S51, simulating starting from the node to be shipped on the transport directed graph; S52, calculating the real-time fitness according to the real-time traffic information of the road section corresponding to the current connecting line to be passed; S53, adding the estimated fitness and the real-time fitness of the same connection line to obtain a score of the connection line; S54, at the intersection of each connection line along the route, based on the score of the connection line, establishing a connection line selection objective function; S55. Select the connecting line corresponding to the maximum target value of the objective function as the optimal connecting line until the node to be received is reached to obtain the optimal route.

[0011] Furthermore, the formula for calculating the real-time fitness in S52 is: , where h r is the real-time fitness, e is a natural constant, x a is the real-time accident number, x t is the real time passing time, x p is the road surface roughness, ω a For x a The weight, ω t For x t The weight, ω p For x p The weight of .

[0012] Furthermore, the objective function selected in S54 is: , , where s i,j is the target value of the jth optional connection line at the i-th intersection, score i,j is the score of the jth optional connection line when passing through the i-th intersection, score i,j,n is the score of the nth optional connection line at the next intersection after the jth optional connection line passing through the ith intersection, i is the number of the intersection, j is the number of the optional connection line at the ith intersection, n is the number of the optional connection line at the next intersection after the optional connection line at the ith intersection, max2 is the maximum score, M is the number of optional connections at the ith intersection, and max1 is the connection line corresponding to the maximum target value among the M target values.

[0013] Furthermore, S6 is specifically as follows: obtaining the logistics risk according to the gap between the time attribute value, economic attribute value, risk attribute value and supply attribute value on each connecting line on the optimal route and the set time target value, economic target value, risk target value and supply target value.

[0014] The beneficial effects of the present invention are as follows: the present invention realizes a comprehensive representation of a complex logistics network by constructing a transport directed graph of a multi-source heterogeneous fusion database. Secondly, the introduction of multidimensional attribute values ​​(time, economy, risk and supply) to comprehensively evaluate the connection lines greatly improves the accuracy and adaptability of path selection. By setting the importance of different attributes and calculating the weights of the connection lines, it is possible to flexibly respond to the needs of different logistics scenarios. In particular, the calculation of estimated fitness and the dynamic selection mechanism based on real-time road conditions enable timely response to changes in the logistics environment, selection of the optimal connection line, improved accuracy of path selection, and enhanced real-time optimization capabilities of the logistics process. Finally, by conducting risk assessments on each connection line experienced during the simulation process, comprehensive risk insights are provided for logistics management, which helps prevent and reduce potential logistics problems.

[0015] In the present invention, the importance of time, economy, risk and replenishment is set according to demand in step S3, so as to achieve adaptive adjustment of the weight of each connection line according to the transportation demand. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The figure is a flow chart of a cross-risk control method based on cross-regional logistics data. DETAILED DESCRIPTION

[0017] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0018] like Figure 1 As shown, a cross-risk control method based on cross-regional logistics data includes the following steps: S1. Construct a transportation directed graph based on a multi-source heterogeneous fusion database; S2. Generate a multidimensional attribute value for each connection line in the transport directed graph, wherein the multidimensional attribute value includes: a time attribute value, an economic attribute value, a risk attribute value, and a supply attribute value; S3. Set the importance of time, economy, risk and supply, and calculate the weight of the connection line based on the multi-dimensional attribute value; S4. Calculate the estimated fitness of the connection line according to the weight of the connection line; S5. On the transport directed graph, it is simulated that starting from the node to be shipped, based on the estimated fitness of the connection line and the real-time road conditions, at the intersection of each connection line, a connection line selection objective function is established to select the optimal connection line until the node to be received is reached, and the optimal route is obtained; S6. Evaluate logistics risks based on the connecting lines on the optimal route.

[0019] In this embodiment, the importance of time, economy, risk and replenishment is set according to demand in step S3, so as to achieve adaptive adjustment of the weight of each connection line according to the transportation demand.

[0020] Refueling stations can be: gas stations, charging stations and rest stops.

[0021] The multi-source heterogeneous fusion database is a comprehensive data management system that integrates data sets from multiple different sources with different structures and formats. Therefore, in the present invention, the multi-source heterogeneous fusion database is a database of various forms of logistics information.

[0022] Specifically, S1 is as follows: each receiving location in the multi-source heterogeneous fusion database is used as a receiving node, each shipping location is used as a shipping node, and each path between the receiving location and the receiving node is mapped into a connecting line to obtain a transport directed graph.

[0023] The time attribute value in S2 is the normalized value of the truck passing time of each connecting line; The economic attribute value is a normalized value of the economic cost consumed on each connection line; The risk attribute value is a normalized value of the weighted values ​​of cargo damage risk, theft risk, traffic accident risk and natural disaster risk on each connection line; The supply attribute value is the normalized value of the number of supply stations on each connecting line.

[0024] In this embodiment, the time attribute value is the normalized value of the truck passing time of each connecting line. Each connecting line is a route. Therefore, the truck passing time can be obtained by the length of the route and the average vehicle speed.

[0025] In this embodiment, the supply attribute value is the normalized value of the number of supply stations on each connecting line. The supply stations can be gas stations, charging stations, etc. For electric vehicle transportation, the number of charging stations on the route is counted and normalized to obtain the supply attribute value. For oil truck transportation, the number of gas stations on the route is counted and normalized to obtain the supply attribute value.

[0026] In this embodiment, the economic attribute value is the normalized value of the economic cost consumed on each connecting line. The consumed economic cost includes: fuel consumption cost, road loss cost, line charging cost, etc. The cost generated by passing through the connecting line is added together to obtain the consumed economic cost.

[0027] In this embodiment, the risk attribute value is a normalized value of the weighted values ​​of cargo damage risk, theft risk, traffic accident risk and natural disaster risk on each connecting line. The cargo damage risk is the proportion of cargo damage when the route is passed in history, the theft risk is the proportion of cargo theft when the route is passed in history, the traffic accident risk is the proportion of traffic accidents when the route is passed in history, and the natural disaster risk is the proportion of natural disaster risks when the route is passed in history. These proportions can be set through experience or obtained by searching historical data.

[0028] The S3 comprises the following sub-steps: S31. Set the importance of time, economy, risk and supply, with the importance ranging from 0 to 1; S32. Construct a weight calculation function based on the importance and multi-dimensional attribute values ​​to obtain the weight of the connecting line.

[0029] The weight calculation function in S32 is: , where v is the weight of the connection line, t is the time attribute value of the connection line, E is the economic attribute value of the connection line, r is the risk attribute value of the connection line, s is the supply attribute value of the connection line, θ t is the importance of the time attribute value of the connection line, θ E is the importance of the economic attribute value of the connecting line, θ r is the importance of the risk attribute value of the connecting line, θ s is the importance of the replenishment attribute value of the connecting line.

[0030] By setting the importance of the four dimensions of time, economy, risk and supply, the present invention can flexibly adapt to the needs of different logistics scenarios and make decisions more accurate. Secondly, the weight calculation function comprehensively considers the impact of each attribute value on the importance of the connection line, so that the weight reflects the numerical size of the dimension concerned by transportation.

[0031] The importance of time, economy, risk, and supply all range from 0 to 1.

[0032] The importance should be determined based on the following aspects: a) Specific requirements of transportation tasks: For tasks with high timeliness requirements (such as the transportation of fresh agricultural products), the importance of time should be higher. For cost-sensitive tasks, the importance of economy should be higher. For high-value or fragile goods, the importance of risk should be higher. For long-distance transportation or special vehicles (such as electric vehicles), the importance of supply should be higher. b) Industry characteristics: In cold chain logistics, time and risk are usually more important. In the transportation of bulk commodities, economic factors are more important. c) Seasonal and weather factors: In severe weather conditions, the importance of risk needs to be increased. In hot seasons, for cold chain transportation, the importance of time and supply (such as refrigeration equipment) needs to be increased.

[0033] In the cold chain transportation process, the importance of time, economy, risk and supply can be set to 1, 0.5, 0, 0, reflecting that time is the most critical factor (fresh products have high requirements for timeliness), followed by economic factors (cost control is also very important), and risk and supply are not considered for the time being. It is suitable for cold chain transportation scenarios with short distances, high requirements for timeliness, cost sensitivity, but relatively controllable risks. In the present invention, when t is smaller, E is smaller, r is smaller, and s is larger, the weight v is larger.

[0034] The formula for calculating the estimated fitness of the connecting line in S4 is: , where h e is the estimated fitness of the connection line, α is the proportional coefficient of the weight of the connection line, v is the weight of the connection line, d is the straight-line distance from the node to be received after passing through the connection line, β is the proportional coefficient of the straight-line distance, and D is the maximum straight-line distance.

[0035] Although the weight of each connecting line is obtained through weight calculation, which reflects the matching degree between the transportation demand and each connecting line, not every connecting line can reach the node to be received. Therefore, the distance term (1-d / D) designed in the present invention takes into account the impact of path selection on the overall distribution efficiency and encourages the selection of a path closer to the destination.

[0036] In this embodiment, the proportionality coefficient α and the proportionality coefficient β of the straight-line distance can both be set to 0.5, and the specific values ​​can be adjusted according to needs.

[0037] The proportional coefficients α and β are proportional parameters used to adjust the estimated fitness calculation. The maximum straight-line distance can be the straight-line distance between the shipping point and the receiving point during a transportation process. At the same time, the maximum straight-line distance can also be set as a fixed parameter. The role of the fixed parameter is to normalize d.

[0038] In this embodiment, the proportional coefficient α and the proportional coefficient β of the straight-line distance can be assigned values ​​​​one by one through algorithms such as genetic algorithms and gradient descent methods, until α and β corresponding to the minimum economic cost and the shortest time are found.

[0039] The S5 comprises the following sub-steps: S51, simulating starting from the node to be shipped on the transport directed graph; S52, calculating the real-time fitness according to the real-time traffic information of the road section corresponding to the current connecting line to be passed; S53, adding the estimated fitness and the real-time fitness of the same connection line to obtain a score of the connection line; S54, at the intersection of each connection line along the route, based on the score of the connection line, establishing a connection line selection objective function; S55. Select the connecting line corresponding to the maximum target value of the objective function as the optimal connecting line until the node to be received is reached to obtain the optimal route.

[0040] The present invention simulates starting from the starting point on the transport directed graph, calculates the real-time fitness based on the real-time road condition information, and flexibly responds to the dynamically changing traffic environment. The estimated and real-time fitness are added to obtain the connection line score, which takes into account both the expected effect and the actual situation, making the evaluation more comprehensive and accurate.

[0041] At the intersection (crossing) of each connecting line, a selection objective function is established based on the score, and the connecting line corresponding to the maximum value of the objective function is selected as the optimal connecting line, ensuring that each step is the current optimal one.

[0042] In the present invention, the greater the estimated fitness and the real-time fitness, the more the connecting line meets the requirements, therefore, the connecting line corresponding to the maximum value of the objective function is selected as the optimal connecting line.

[0043] The formula for calculating the real-time fitness in S52 is: , where h r is the real-time fitness, e is a natural constant, x a is the real-time accident number, x t is the real time passing time, x p is the road surface roughness, ω a For x a The weight, ω t For x t The weight, ω p For x p The weight of .

[0044] In this embodiment, ω a ,ω t ,ω p Make specific settings based on experiments or experience.

[0045] In this embodiment, a a ,ω t ,ω p Specific acquisition method: Collect historical transportation data, including real-time accident data x a , real-time passing time x t and road surface roughness x p , get samples, build multiple samples into a training set, manually mark each sample with a label, input the samples in the training set into the formula of real-time fitness, get the real-time fitness of the training process, subtract the real-time fitness of the training process from the label, get the difference, adjust the weight according to the difference (that is, the existing gradient descent method), until the difference is minimum, and the weight adjustment ends, ω a ,ω t ,ω p Get. a ,ω t ,ω p Specific acquisition methods also include: existing genetic algorithms. In the present invention, the smaller the real-time passing time, the smaller the road surface roughness, and the smaller the real-time accident number, the greater the real-time fitness, and therefore, the greater the real-time fitness, the better the real-time situation.

[0046] The objective function selected in S54 is: , , where s i,j is the target value of the jth optional connection line at the i-th intersection, score i,jis the score of the jth optional connection line when passing through the i-th intersection, score i,j,n is the score of the nth optional connection line at the next intersection after the jth optional connection line passing through the ith intersection, i is the number of the intersection, j is the number of the optional connection line at the ith intersection, n is the number of the optional connection line at the next intersection after the optional connection line at the ith intersection, max2 is the maximum score, M is the number of optional connections at the ith intersection, and max1 is the connection line corresponding to the maximum target value among the M target values.

[0047] The present invention combines the current connection line score (score i,j ) and the best possible score for the next step (max2{score i,j,n}), not only taking into account immediate benefits, but also taking into account long-term interests, achieving a balance between local and global. Secondly, the use of the max2 function to select the best score for the next step ensures the foresight and optimization of the decision. Furthermore, the max1 function selects the one with the largest target value from all possible connection lines, ensuring that each step is the best choice at the moment, so that the path selection process can not only adapt to the current situation, but also make the best preparation for the future, greatly improving the efficiency and reliability of the overall logistics path.

[0048] In the present invention, at the i-th intersection, there are M selectable connecting lines. After any connecting line is selected to pass through, at the next intersection, the maximum score is selected, and the scores of the two connecting lines are combined to obtain the target value, thereby selecting the optimal connecting line.

[0049] In this embodiment, the connection line may be selected as a path through which the node to be received can be reached.

[0050] In the present invention, the current connection line score i,j ) consists of estimated fitness and real-time fitness, (score i,j ) score quantifies the quality of a link by combining estimated fitness and real-time fitness. The real-time fitness has a clear calculation formula that takes into account real-time incidents, transit time, and road conditions.

[0051] Specifically, S6 is as follows: obtaining the logistics risk according to the gap between the time attribute value, economic attribute value, risk attribute value and supply attribute value on each connecting line on the optimal route and the set time target value, economic target value, risk target value and supply target value.

[0052] In this embodiment, the first implementation method of step S6 is specifically as follows: adding the time attribute values ​​on each connecting line on the optimal route, subtracting the addition result from the time target value, and obtaining a first difference; adding the economic attribute values ​​on each connecting line on the optimal route, subtracting the addition result from the economic target value, and obtaining a second difference; adding the risk attribute values ​​on each connecting line on the optimal route, subtracting the addition result from the risk target value, and obtaining a third difference; adding the supply attribute values ​​on each connecting line on the optimal route, subtracting the addition result from the supply target value, and obtaining a fourth difference; weighting the first difference, the second difference, the third difference, and the fourth difference to obtain logistics risk. The greater the logistics risk, the greater the risk of deviation from the multidimensional target value.

[0053] In this embodiment, the second implementation method of step S6 is specifically as follows: comparing the data of the corresponding dimension with the target value of the corresponding dimension in turn, and finding the amount that exceeds the target, for example, comparing the time attribute value with the time target value, if it is lower than the time target value, the transportation time is met, and if it is higher than the time target value, the transportation time is not met; comparing the economic attribute value with the economic target value, if it is lower than the economic target value, the transportation economic cost is met, and if it is higher than the economic target value, the transportation economic cost is not met; comparing the risk attribute value with the risk target value, if it is lower than the risk target value, the transportation risk is met, and if it is higher than the risk target value, the transportation risk is not met; comparing the supply attribute value with the supply target value, if it is lower than the supply target value, the transportation supply is not met, and if it is higher than the supply target value, the transportation supply is met; selecting the unsatisfied items and calculating the difference, the larger the difference, the greater the logistics risk.

[0054] The present invention realizes a comprehensive representation of complex logistics networks by constructing a transport directed graph of a multi-source heterogeneous fusion database. Secondly, the multi-dimensional attribute values ​​(time, economy, risk and supply) are introduced to comprehensively evaluate the connection lines, which greatly improves the accuracy and adaptability of path selection. By setting the importance of different attributes and calculating the weights of the connection lines, it is possible to flexibly respond to the needs of different logistics scenarios. In particular, the calculation of estimated fitness and the dynamic selection mechanism based on real-time road conditions enable timely response to changes in the logistics environment, selection of the optimal connection line, improvement of the accuracy of path selection, and enhancement of the real-time optimization capability of the logistics process. Finally, by conducting risk assessment on each connection line experienced during the simulation process, comprehensive risk insight is provided for logistics management.

[0055] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A cross-risk control method based on cross-regional logistics data, characterized in that: The following steps are included: S1. Construct a transportation directed graph based on a multi-source heterogeneous fusion database; S2. generating a multidimensional attribute value for each connecting line in the transport directed graph, wherein the multidimensional attribute value includes: a time attribute value, an economic attribute value, a risk attribute value, and a supply attribute value; S3, set the importance of time, economy, risk and supply, and calculate the weight of the connection line based on the multi-dimensional attribute value; S4. Calculate the estimated fitness of the connection line according to the weight of the connection line; S5. On the transport directed graph, it is simulated to start from the node to be shipped, and based on the estimated fitness of the connection line and the real-time road conditions, at the intersection of each connection line, a connection line selection objective function is established to select the optimal connection line until the node to be received is reached, and the optimal route is obtained; S6. Evaluate logistics risks based on the connecting lines on the optimal route.

2. The cross-risk control method based on cross-regional logistics data according to claim 1 is characterized in that: Specifically, S1 is as follows: each receiving location in the multi-source heterogeneous fusion database is used as a receiving node, each shipping location is used as a shipping node, and each path between the receiving location and the receiving node is mapped into a connecting line to obtain a transport directed graph.

3. The cross-risk control method based on cross-regional logistics data according to claim 1 is characterized in that: The time attribute value in S2 is the normalized value of the truck passing time of each connecting line; The economic attribute value is a normalized value of the economic cost consumed on each connection line; The risk attribute value is a normalized value of the weighted values ​​of cargo damage risk, theft risk, traffic accident risk and natural disaster risk on each connection line; The supply attribute value is the normalized value of the number of supply stations on each connecting line.

4. The cross-risk control method based on cross-regional logistics data according to claim 1 is characterized in that: The S3 comprises the following sub-steps: S31. Set the importance of time, economy, risk and supply, with the importance ranging from 0 to 1; S32. Construct a weight calculation function based on the importance and multi-dimensional attribute values ​​to obtain the weight of the connecting line.

5. The cross-risk control method based on cross-regional logistics data according to claim 4 is characterized in that: The weight calculation function in S32 is: , where v is the weight of the connection line, t is the time attribute value of the connection line, E is the economic attribute value of the connection line, r is the risk attribute value of the connection line, s is the supply attribute value of the connection line, θ t is the importance of the time attribute value of the connection line, θ E is the importance of the economic attribute value of the connecting line, θ r is the importance of the risk attribute value of the connecting line, θ s is the importance of the replenishment attribute value of the connecting line.

6. The cross-risk control method based on cross-regional logistics data according to claim 1 is characterized in that: The formula for calculating the estimated fitness of the connecting line in S4 is: , where h e is the estimated fitness of the connection line, α is the proportional coefficient of the weight of the connection line, v is the weight of the connection line, d is the straight-line distance from the node to be received after passing through the connection line, β is the proportional coefficient of the straight-line distance, and D is the maximum straight-line distance.

7. The cross-risk control method based on cross-regional logistics data according to claim 1 is characterized in that: The S5 comprises the following sub-steps: S51, simulating starting from the node to be shipped on the transport directed graph; S52, calculating the real-time fitness according to the real-time traffic information of the road section corresponding to the current connecting line to be passed; S53, adding the estimated fitness and the real-time fitness of the same connection line to obtain a score of the connection line; S54, at the intersection of each connection line along the route, based on the score of the connection line, establishing a connection line selection objective function; S55. Select the connecting line corresponding to the maximum target value of the objective function as the optimal connecting line until the node to be received is reached to obtain the optimal route.

8. The cross-risk control method based on cross-regional logistics data according to claim 7 is characterized in that: The formula for calculating the real-time fitness in S52 is: , where h r is the real-time fitness, e is a natural constant, x a is the real-time accident number, x t is the real time passing time, x p is the road surface roughness, ω a For x a The weight, ω t For x t The weight, ω p For x p The weight of .

9. The cross-risk control method based on cross-regional logistics data according to claim 7 is characterized in that: The objective function selected in S54 is: , , where s i,j is the target value of the jth optional connection line at the i-th intersection, score i,j is the score of the jth optional connection line when passing through the i-th intersection, score i,j,n is the score of the nth optional connection line at the next intersection after the jth optional connection line passing through the ith intersection, i is the number of the intersection, j is the number of the optional connection line at the ith intersection, n is the number of the optional connection line at the next intersection after the optional connection line at the ith intersection, max2 is the maximum score, M is the number of optional connections at the ith intersection, and max1 is the connection line corresponding to the maximum target value among the M target values.

10. The cross-risk control method based on cross-regional logistics data according to claim 1 is characterized in that: Specifically, S6 is as follows: obtaining the logistics risk according to the gap between the time attribute value, economic attribute value, risk attribute value and supply attribute value on each connecting line on the optimal route and the set time target value, economic target value, risk target value and supply target value.

Images