Multi-target agricultural product logistics distribution path optimization method based on ant colony algorithm

Abstract

The invention discloses a multi-target agricultural product logistics distribution path optimization method based on an ant colony algorithm, and particularly relates to the technical field of distribution path optimization, and the method comprises the steps: constructing an initial distribution graph with time window constraint based on the geographic coordinates of a distribution center and a delivery point, and monitoring the distribution sequence state of agricultural products in real time in the distribution process, thereby achieving the optimal distribution path optimization of the agricultural products. The method can accurately identify the upside-down distribution sequence caused by emergency, calculates the frequent opening and closing influence coefficient of the compartment door according to the obtained opening and closing information of the compartment door when the distribution sequence is in an abnormal state, and calculates the frequent starting and stopping influence coefficient of the cold-chain equipment according to the starting and stopping information of the cold-chain equipment. And constructing an agricultural product corrosion rate influence evaluation model, outputting an agricultural product corrosion rate influence evaluation index, evaluating the influence rise degree of the cold chain environment on the agricultural product corrosion rate, and improving the existing ant colony algorithm based on the agricultural product corrosion rate influence evaluation index to optimize the logistics distribution path of the agricultural product.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution path optimization, and more specifically, to a multi-objective agricultural product logistics distribution path optimization method based on an ant colony algorithm. Background Art

[0002] In cold chain logistics and distribution of agricultural products, effective route planning and optimized loading and unloading sequences are crucial for ensuring the quality of agricultural products. However, during actual transportation, emergencies (such as road closures, sudden order changes, and adjusted customer demands) may necessitate adjustments to the planned delivery sequence, resulting in a reversal of the delivery order. This change can trigger a series of chain reactions, impacting logistics efficiency and the stability of the cold chain environment. First, agricultural products are typically loaded according to the "first-out, last-in" principle, with cargo near the vehicle door prioritized to reduce loading and unloading complexity. However, when the delivery sequence changes, if the target agricultural products are placed inside the vehicle, the outer layers of cargo must be unloaded first, the target cargo removed, and finally the outer layers reloaded. This frequent opening and closing of the vehicle doors disrupts the cold chain environment, causing temperature fluctuations and increasing the risk of spoilage. Second, cold chain logistics typically operates within strict time windows, requiring delivery vehicles to complete deliveries within a specified timeframe. To compensate for the extended loading and unloading time caused by sequence adjustments, vehicles may be forced to accelerate to ensure on-time arrival. However, speeding increases engine load, impacting the stable operation of cold chain equipment. When refrigeration equipment is under high load for a long time, it may trigger an automatic protective shutdown, causing the temperature inside the carriage to rise rapidly. This exposes agricultural products to high temperatures, accelerating microbial growth, leading to quality degradation, and even irreversible food safety issues. Traditional path optimization algorithms focus primarily on the shortest path or lowest cost, but fail to fully consider practical factors such as delivery sequence changes, cold chain equipment load management, and emergency response. Therefore, the impact of frequent opening and closing of carriage doors and the frequent start-up and shutdown of cold chain equipment caused by delivery sequence changes within time window constraints on the spoilage rate must be comprehensively considered during the cold chain logistics optimization process to achieve the optimal delivery plan, ensure delivery efficiency, and maintain the stability of the cold chain environment. Summary of the Invention

[0003] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a multi-objective agricultural product logistics distribution path optimization method based on an ant colony algorithm to solve the problems raised in the above-mentioned background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] The multi-objective agricultural product logistics distribution path optimization method based on ant colony algorithm includes the following steps:

[0006] Step S1: constructing an initial distribution map with time window constraints based on the geographic coordinates of the distribution center and the delivery point, distributing the agricultural products according to the initial distribution map, and monitoring the distribution sequence status of the agricultural products, wherein the distribution sequence status of the agricultural products includes a normal distribution sequence status and an abnormal distribution sequence status;

[0007] Step S2: When the delivery sequence of the agricultural products is in an abnormal delivery sequence, the opening and closing information of the carriage doors is obtained during the delivery of the agricultural products, and the frequent opening and closing influence coefficient of the carriage doors is calculated based on the opening and closing information of the carriage doors to evaluate the impact of the frequent opening and closing of the carriage doors on the corrosion rate of the agricultural products;

[0008] Step S3: When the delivery order status of the agricultural products is in an abnormal delivery order state, the start and stop information of the cold chain equipment is obtained during the delivery process of the agricultural products, and the frequent start and stop impact coefficient of the cold chain equipment is calculated based on the start and stop information of the cold chain equipment to evaluate the impact of the frequent start and stop of the cold chain equipment on the corrosion rate of the agricultural products;

[0009] Step S4: constructing an agricultural product corrosion rate impact assessment model based on the impact coefficient of frequent door opening and closing and the impact coefficient of frequent start and stop of cold chain equipment, outputting an agricultural product corrosion rate impact assessment index, and assessing the degree of increase in the impact of agricultural product corrosion rate;

[0010] Step S5: improving the existing ant colony algorithm according to the degree of increase in the impact of the agricultural product corrosion rate to optimize the logistics distribution path of the agricultural products.

[0011] In a preferred embodiment, the logic for obtaining the frequent opening and closing influence coefficient of the carriage door is as follows:

[0012] During the agricultural product distribution process, the door opening and closing information is obtained, including the number of door openings, the duration of a single door opening, the temperature inside and outside the door, and the door opening area. The heat exchange between the door and the outside world is calculated when the door is open: Qj = dl*mj*ΔT*Topen, where Qj represents the heat exchange between the door and the outside world, dl represents the convection heat transfer coefficient, mj represents the door opening area, ΔT represents the temperature difference between the inside and outside of the door, and Topen represents the duration of the door opening. The temperature fluctuation is calculated each time the door is opened: Where twi represents the temperature fluctuation when the door is opened, jl represents the air mass in the compartment, jl = md*tj, where md represents the air density and tj represents the interior volume of the compartment;

[0013] The temperature fluctuation data set KB is constructed by the temperature fluctuation each time the door is opened. i}={twi1,twi2,...,twi I}, where twi irepresents the temperature fluctuation when the i-th door is opened, i = {1, 2, ..., I}, I is the total number of times the door is opened;

[0014] Calculate the correlation coefficient between the temperature fluctuation of the i-th compartment door opening and the temperature fluctuation of the remaining compartment door openings in the temperature fluctuation data set: Where xg represents the correlation coefficient between the temperature fluctuation of the i-th door opening and the temperature fluctuation of the j-th door opening in the temperature fluctuation data set, twi j represents the temperature fluctuation of the remaining j-th compartment door opening in the temperature fluctuation data set, and σt represents the standard deviation of the temperature fluctuation of all compartment door openings in the temperature fluctuation data set;

[0015] Compare the correlation coefficient with a preset correlation coefficient threshold. If the correlation coefficient is greater than the correlation coefficient threshold, add the temperature fluctuation of the j-th door opening to the similar neighborhood set of the temperature fluctuation of the i-th door opening.

[0016] Each door opening is defined as a node, and the correlation coefficient between two door opening and closing events is used as the edge weight to construct the door opening and closing-temperature fluctuation propagation network;

[0017] Calculate the node influence as follows:

[0018] Step A1: Initialize the initial node influence of all nodes: PR 0 (i) represents the node influence at iteration 0, i = {1, 2, ..., I}, I is the total number of door openings;

[0019] Step A2: For each node i, calculate the new node influence in this round based on the node influence in the previous round: PR t+1 (i) represents the node influence at the t+1th iteration, N(j) represents the similar neighborhood set of node i, deg represents the total number of nodes in the similar neighborhood set, PR t (j) represents the node influence at the tth iteration, and d represents the damping factor;

[0020] Step A3: Repeat step A2 to traverse all nodes and continuously update the node influence. The difference between the node influences of two consecutive rounds is compared with the preset convergence threshold. If the difference between the node influences of two consecutive rounds is less than the convergence threshold, the iteration is stopped.

[0021] Calculate the influence coefficient of frequent opening and closing of carriage doors: Where Cdoor represents the influence coefficient of frequent door opening and closing, and PR(i) represents the node influence of the i-th door opening event.

[0022] In a preferred embodiment, the logic for obtaining the influence coefficient of frequent start-stop of cold chain equipment is as follows:

[0023] Obtain the compartment temperature each time the cold chain equipment is started, and compare the compartment temperature each time the cold chain equipment is started with the refrigeration temperature of the agricultural products. If the compartment temperature at the time the cold chain equipment is started is higher than the refrigeration temperature of the agricultural products, add the compartment temperature at the time of the cold chain equipment startup to the temperature data set;

[0024] Calculate the temperature recovery time for the carriage temperature to return to the agricultural product refrigeration temperature in the temperature data set: τrec=min{t-ton|T(t)≤Tset,T(t)>Tset+δ}, where τrec represents the temperature recovery time, t-ton represents the time interval calculated from the cold chain equipment startup time ton, T(t) represents the carriage temperature at time t, Tset represents the refrigeration temperature of the agricultural product, and δ is the temperature offset value, indicating that the carriage temperature must exceed this value before cooling is triggered;

[0025] Calculate the mean temperature recovery time for the carriage temperature to return to the agricultural product refrigeration temperature in the temperature data set: Where jz represents the mean temperature recovery time of the carriage temperature in the temperature data set to return to the agricultural product refrigeration temperature, τrec g represents the temperature recovery time for the g-th compartment temperature in the temperature data set to return to the agricultural product refrigeration temperature, g = {1, 2, ..., G}, where G is a positive integer;

[0026] Calculate the standard deviation of the temperature recovery time for the carriage temperature to return to the agricultural product refrigeration temperature in the temperature data set: Where bz represents the standard deviation of the temperature recovery time for the compartment temperature to return to the agricultural product refrigeration temperature in the temperature data set;

[0027] Calculate the impact coefficient of frequent start and stop of cold chain equipment: Ccool represents the impact coefficient of frequent start and stop of cold chain equipment.

[0028] In a preferred embodiment, an agricultural product corrosion rate impact assessment model is constructed based on the impact coefficient of frequent opening and closing of the carriage door and the impact coefficient of frequent start and stop of the cold chain equipment, and the agricultural product corrosion rate impact assessment index CPI is output. The model is based on the following formula: CPI=a1*Cdoor+a2*Ccool, where a1 and a2 represent the preset proportional coefficients of the impact coefficient of frequent opening and closing of the carriage door and the impact coefficient of frequent start and stop of the cold chain equipment, respectively, and a1 and a2 are both greater than 0.

[0029] In a preferred embodiment, the agricultural product corrosion rate impact assessment index is compared with a preset agricultural product corrosion rate impact assessment index threshold to assess the degree of increase in the impact of the agricultural product corrosion rate, as follows:

[0030] If the agricultural product corrosion rate impact assessment index is greater than the agricultural product corrosion rate impact assessment index threshold, a high impact rising degree signal is generated; if the agricultural product corrosion rate impact assessment index is less than or equal to the agricultural product corrosion rate impact assessment index threshold, a low impact rising degree signal is generated.

[0031] In a preferred embodiment, when a high impact rising degree signal is generated, the existing ant colony algorithm is improved to optimize the logistics distribution path of agricultural products, specifically as follows:

[0032] Step B1, initializing the ant colony: setting the number of ants and initializing pheromones;

[0033] Step B2, path construction optimization: Each ant starts from the current delivery point and determines the probability of choosing the next stop based on the pheromone concentration and heuristic factor;

[0034] Step B3, pheromone update: after each ant selects a path, the pheromone is updated immediately;

[0035] Step B4: preset the maximum number of iterations, repeat steps B2 and B3 in a loop, and output the delivery path after the iteration stops.

[0036] The technical effects and advantages of the present invention are as follows:

[0037] 1. The present invention constructs an initial distribution map with time window constraints based on the geographic coordinates of the distribution center and the delivery point, and monitors the distribution sequence of agricultural products in real time during the distribution process. It can accurately identify the reversal of the distribution sequence caused by emergency situations. When the distribution sequence is in an abnormal state, the frequent opening and closing influence coefficient of the carriage door is calculated based on the acquired carriage door opening and closing information. At the same time, the frequent opening and closing influence coefficient of the cold chain equipment is calculated using the start and stop information of the cold chain equipment. Then, an agricultural product corrosion rate impact assessment model is constructed and an agricultural product corrosion rate impact assessment index is output, thereby quantitatively assessing the degree of increase in the impact of the cold chain environment on the corrosion rate of agricultural products. Based on the agricultural product corrosion rate impact assessment index, the existing The ant colony algorithm is improved by comprehensively considering multiple objective factors such as path distance, frequent opening and closing of carriage doors, and start-up and shutdown of cold chain equipment in the path selection process, optimizing the pheromone update rules, ensuring that the selected distribution path can minimize temperature fluctuations and equipment load while meeting the time window constraints, thereby reducing the risk of agricultural product corrosion, improving the stability of cold chain transportation and logistics distribution efficiency. It not only realizes real-time dynamic monitoring of the cold chain logistics environment, but also provides a precise control basis for the cold chain transportation of agricultural products through multi-objective optimization decision-making, significantly improving the storage quality and safety of agricultural products during transportation, while reducing energy consumption and equipment loss caused by unstable temperature control. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings;

[0039] Figure 1 Flowchart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0041] Example: Figure 1 The present invention provides a multi-objective agricultural product logistics distribution path optimization method based on ant colony algorithm, which includes the following steps:

[0042] Step S1: constructing an initial distribution map with time window constraints based on the geographic coordinates of the distribution center and the delivery point, distributing the agricultural products according to the initial distribution map, and monitoring the distribution sequence status of the agricultural products, wherein the distribution sequence status of the agricultural products includes a normal distribution sequence status and an abnormal distribution sequence status;

[0043] Step S2: When the delivery sequence of the agricultural products is in an abnormal delivery sequence, the opening and closing information of the carriage doors is obtained during the delivery of the agricultural products, and the frequent opening and closing influence coefficient of the carriage doors is calculated based on the opening and closing information of the carriage doors to evaluate the impact of the frequent opening and closing of the carriage doors on the corrosion rate of the agricultural products;

[0044] Step S3: When the delivery order status of the agricultural products is in an abnormal delivery order state, the start and stop information of the cold chain equipment is obtained during the delivery process of the agricultural products, and the frequent start and stop impact coefficient of the cold chain equipment is calculated based on the start and stop information of the cold chain equipment to evaluate the impact of the frequent start and stop of the cold chain equipment on the corrosion rate of the agricultural products;

[0045] Step S4: constructing an agricultural product corrosion rate impact assessment model based on the impact coefficient of frequent door opening and closing and the impact coefficient of frequent start and stop of cold chain equipment, outputting an agricultural product corrosion rate impact assessment index, and assessing the degree of increase in the impact of agricultural product corrosion rate;

[0046] Step S5, improving the existing ant colony algorithm according to the degree of increase in the impact of the agricultural product corrosion rate to optimize the logistics distribution path of the agricultural products;

[0047] In step S1, an initial delivery graph G(V,E) with time window constraints is constructed based on the geographic coordinates of the delivery center and the delivery point, where V is a set of nodes, including the geographic coordinates of the delivery center and the delivery point, and E is a set of paths, including the delivery paths between nodes and delivery attribute information, including delivery distance, estimated travel time, and time window constraints (the time range within which customers can receive goods).

[0048] Determine the order of loading agricultural products based on the initial delivery map, following the "first out, last in" principle, meaning that the first goods to be delivered are placed near the carriage door to reduce unnecessary loading and unloading operations;

[0049] Distribute agricultural products according to the initial distribution map. During the distribution process, continuously monitor the distribution sequence of agricultural products, including:

[0050] Normal delivery sequence: The vehicle delivers according to the planned route, and the delivery sequence of each delivery point remains unchanged;

[0051] Abnormal delivery sequence status: The vehicle cannot deliver according to the planned route due to reasons such as road closures or sudden order changes, and the delivery sequence of each delivery point is changed;

[0052] Step S2: When the delivery sequence of the agricultural products is in an abnormal delivery sequence, the opening and closing information of the carriage doors is obtained during the delivery of the agricultural products, and the frequent opening and closing influence coefficient of the carriage doors is calculated based on the opening and closing information of the carriage doors to evaluate the impact of the frequent opening and closing of the carriage doors on the corrosion rate of the agricultural products;

[0053] In the present invention, the frequent opening and closing influence coefficient of the carriage door is used to measure the influence of the frequent opening and closing of the carriage door on the corrosion rate of agricultural products. A larger frequent opening and closing influence coefficient of the carriage door indicates that the carriage door is opened many times and for a long time, resulting in greater damage to the cold chain environment and drastic temperature fluctuations inside the carriage, thereby accelerating the spoilage of agricultural products. A smaller frequent opening and closing influence coefficient of the carriage door indicates that the carriage door is opened less frequently or for a shorter opening time, the cold chain environment is relatively stable, and the agricultural products are less affected. By evaluating the degree of increase in the influence of the corrosion rate of agricultural products based on the frequent opening and closing influence coefficient of the carriage door, the impact of temperature fluctuations on the quality of agricultural products can be accurately evaluated, providing a scientific basis for optimizing logistics distribution routes, thereby effectively reducing the risk of agricultural product corruption and ensuring distribution quality.

[0054] The logic for obtaining the influence coefficient of frequent door opening and closing is as follows:

[0055] During the agricultural product distribution process, the door opening and closing information is obtained, including the number of door openings, the duration of a single door opening, the temperature inside and outside the door, and the door opening area. The heat exchange between the door and the outside world is calculated when the door is open: Qj = dl*mj*ΔT*Topen, where Qj represents the heat exchange between the door and the outside world, dl represents the convection heat transfer coefficient, mj represents the door opening area, ΔT represents the temperature difference between the inside and outside of the door, and Topen represents the duration of the door opening. The temperature fluctuation is calculated each time the door is opened: Where twi represents the temperature fluctuation when the door is opened, jl represents the air mass in the compartment, jl = md*tj, where md represents the air density and tj represents the interior volume of the compartment;

[0056] It should be noted that the convective heat transfer coefficient can be calculated using the correlation formula between the Greishev number (Gr) and the Nusselt number (Nu): dl = C1*(Gr*Nu) n1 *k1 / LG, where Gr is the Greishev number, Nu is the Nusselt number, C1 and n1 are empirical constants, C1 = 0.59, n1 = 1 / 4, k1 is the thermal conductivity of air, and LG is the height of the car door;

[0057] The temperature fluctuation data set KB is constructed by the temperature fluctuation each time the door is opened. i}={twi1,twi2,...,twi I}, where twi i represents the temperature fluctuation when the i-th door is opened, i = {1, 2, ..., I}, I is the total number of times the door is opened;

[0058] Calculate the correlation coefficient between the temperature fluctuation of the i-th compartment door opening and the temperature fluctuation of the remaining compartment door openings in the temperature fluctuation data set: Where xg represents the correlation coefficient between the temperature fluctuation of the i-th door opening and the temperature fluctuation of the j-th door opening in the temperature fluctuation data set, twi j represents the temperature fluctuation of the remaining j-th compartment door opening in the temperature fluctuation data set, and σt represents the standard deviation of the temperature fluctuation of all compartment door openings in the temperature fluctuation data set;

[0059] It should be noted that the correlation coefficient is used to measure the relationship between temperature fluctuations between two compartment door opening events;

[0060] Compare the correlation coefficient with the preset correlation coefficient threshold. If the correlation coefficient is greater than the correlation coefficient threshold, it means that the temperature fluctuation patterns caused by the two door openings are similar and have a strong correlation. The temperature fluctuation caused by the j-th door opening is added to the similar neighborhood set of the temperature fluctuation caused by the i-th door opening.

[0061] Each door opening is defined as a node, and the correlation coefficient between two door opening and closing events is used as the edge weight to construct the door opening and closing-temperature fluctuation propagation network;

[0062] Calculate the node influence as follows:

[0063] Step A1: Initialize the initial node influence of all nodes: PR 0 (i) represents the node influence at iteration 0, i = {1, 2, ..., I}, I is the total number of door openings;

[0064] It should be noted that the node influence reflects the global influence of the switch event;

[0065] Step A2: For each node i, calculate the new node influence in this round based on the node influence in the previous round: PR t+1 (i) represents the node influence at the t+1th iteration, N(j) represents the similar neighborhood set of node i, deg represents the total number of nodes in the similar neighborhood set, PR t (j) represents the node influence at the tth iteration, and d represents the damping factor (take d = 0.85);

[0066] Step A3: Repeat step A2 to traverse all nodes and continuously update the node influence. The difference between the node influences of two consecutive rounds is compared with the preset convergence threshold. If the difference between the node influences of two consecutive rounds is less than the convergence threshold, the iteration is stopped.

[0067] Calculate the influence coefficient of frequent opening and closing of carriage doors: Where Cdoor represents the influence coefficient of frequent door opening and closing, and PR(i) represents the node influence of the i-th door opening event;

[0068] It should be noted that the above formulas are all dimensionless and numerical calculations. Common dimensionless methods include Min-Max normalization and Z-Score normalization, which will not be described here.

[0069] During the agricultural product delivery process, vehicle door opening and closing information is first collected, including the number of door openings, the duration of each opening, the temperature inside and outside the vehicle, and the door opening area. This information is used to analyze the impact of door openings on temperature fluctuations. Next, based on the heat exchange between the vehicle and the outside world when the door is open, heat loss within the vehicle is calculated. Combined with internal air quality parameters, the temperature fluctuation caused by each door opening is further calculated. The temperature fluctuation data resulting from all door openings is collected to construct a temperature fluctuation data set. The correlation coefficient between the temperature fluctuation for each door opening and the temperature fluctuations for other door openings in the data set is calculated to measure the similarity of the temperature changes caused by different door openings. The correlation coefficient is then compared with a preset correlation coefficient threshold. If the correlation coefficient exceeds the threshold, the temperature fluctuation patterns of the two door openings are considered similar, indicating a strong correlation between their impacts on vehicle temperature. These two door openings are then included in the corresponding similar neighborhood set. Next, each door opening is defined as a node, and the calculated correlation coefficient is used as the edge weight to construct a door opening and closing-temperature fluctuation propagation network to reveal the temperature impact relationship between door opening and closing events. To further analyze the influence of each door opening event in the overall network, the initial node influence of all nodes is initialized, and an iterative method is used to calculate the final influence of each node. Specifically, during each iteration, the influence of each node is updated based on the influence of the nodes in its similar neighborhood set until the difference in node influence between two consecutive rounds is less than the convergence threshold, thus ensuring the stability and accuracy of the calculation. Finally, based on the influence of each node, the influence coefficient of frequent door opening and closing is calculated. This influence coefficient is obtained by accumulating the node influence of all door opening events. It reflects the comprehensive impact of frequent door opening and closing on internal temperature fluctuations in the carriage, providing a quantitative indicator for the storage and transportation environment assessment of agricultural products.

[0070] Step S3: When the delivery order status of the agricultural products is in an abnormal delivery order state, the start and stop information of the cold chain equipment is obtained during the delivery process of the agricultural products, and the frequent start and stop impact coefficient of the cold chain equipment is calculated based on the start and stop information of the cold chain equipment to evaluate the impact of the frequent start and stop of the cold chain equipment on the corrosion rate of the agricultural products;

[0071] In the present invention, the frequent start-stop impact coefficient of cold chain equipment is used to measure the degree of influence of the frequent opening and closing of carriage doors on the corrosion rate of agricultural products. A larger frequent start-stop impact coefficient of cold chain equipment indicates that the cold chain equipment is started and stopped more frequently due to the frequent opening and closing of carriage doors, and the temperature fluctuations inside the carriage are aggravated, causing the agricultural products to be in an unstable temperature state for a long time, thereby accelerating the increase in corrosion rate; a smaller frequent start-stop impact coefficient of cold chain equipment indicates that the start-stop of the cold chain equipment is relatively stable, the temperature environment inside the carriage is relatively stable, and the agricultural products are less affected by temperature fluctuations, thereby reducing the risk of increased corrosion rate. Evaluating the degree of influence of the corrosion rate of agricultural products based on the frequent start-stop impact coefficient of cold chain equipment can accurately quantify the impact of the operating status of cold chain equipment on the storage and transportation quality of agricultural products, and provide a scientific basis for optimizing the start-stop strategy of cold chain equipment and reducing the corrosion rate of agricultural products.

[0072] The logic for obtaining the impact coefficient of frequent start-up and shutdown of cold chain equipment is as follows:

[0073] Obtain the compartment temperature each time the cold chain equipment is started, and compare the compartment temperature each time the cold chain equipment is started with the refrigeration temperature of the agricultural products. If the compartment temperature at the time the cold chain equipment is started is higher than the refrigeration temperature of the agricultural products, add the compartment temperature at the time of the cold chain equipment startup to the temperature data set;

[0074] Calculate the temperature recovery time for the carriage temperature to return to the agricultural product refrigeration temperature in the temperature data set: τrec=min{t-ton|T(t)≤Tset,T(t)>Tset+δ}, where τrec represents the temperature recovery time, t-ton represents the time interval calculated from the cold chain equipment startup time ton, T(t) represents the carriage temperature at time t, Tset represents the refrigeration temperature of the agricultural product, and δ is the temperature offset value, indicating that the carriage temperature must exceed this value before cooling is triggered;

[0075] It should be noted that min{t-ton|T(t)≤Tset,T(t)>Tset+δ} means finding the first moment when the compartment temperature drops to the refrigeration temperature of agricultural products, that is, T(t)≤Tset; the larger τrec is, the more unstable the cold chain equipment control is, and the agricultural products are kept at a non-optimal temperature for a long time;

[0076] Calculate the mean temperature recovery time for the carriage temperature to return to the agricultural product refrigeration temperature in the temperature data set: Where jz represents the mean temperature recovery time of the carriage temperature in the temperature data set to return to the agricultural product refrigeration temperature, τrec g represents the temperature recovery time for the g-th compartment temperature in the temperature data set to return to the agricultural product refrigeration temperature, g = {1, 2, ..., G}, where G is a positive integer;

[0077] Calculate the standard deviation of the temperature recovery time for the carriage temperature to return to the agricultural product refrigeration temperature in the temperature data set: Where bz represents the standard deviation of the temperature recovery time for the compartment temperature to return to the agricultural product refrigeration temperature in the temperature data set;

[0078] Calculate the impact coefficient of frequent start and stop of cold chain equipment: Where Ccool represents the impact coefficient of frequent start and stop of cold chain equipment;

[0079] It should be noted that the above formulas are all dimensionless and numerical calculations. Common dimensionless methods include Min-Max normalization and Z-Score normalization, which will not be described here.

[0080] In the process of obtaining the impact coefficient of frequent start-up and shutdown of cold chain equipment, it is first necessary to record the startup time of each cold chain equipment and simultaneously collect the compartment temperature data at this time. In order to evaluate the impact of the startup of cold chain equipment on the refrigerated environment of agricultural products, the compartment temperature at each startup of the cold chain equipment is compared with the refrigerated temperature of the agricultural products. If the compartment temperature at startup is higher than the refrigerated temperature of the agricultural products, the temperature data is included in the temperature data set for subsequent analysis of temperature recovery. Next, for the compartment temperature data in the temperature data set, the recovery time required for the compartment temperature to drop from above the refrigerated temperature of agricultural products to the refrigerated temperature is calculated. Specifically, starting from the moment the cold chain equipment is started, the change of the compartment temperature over time is tracked until the compartment temperature returns to the refrigerated temperature range of agricultural products. This time interval is recorded as the temperature recovery time. In order to more comprehensively characterize the impact of cold chain equipment on the temperature recovery process, all temperature recovery times in the temperature data set are statistically analyzed, and their mean is calculated to measure the average time it takes for the cold chain equipment to recover temperature during the entire operation process. At the same time, their standard deviation is calculated to evaluate the stability of the temperature recovery time. Finally, the impact coefficient of frequent cold chain equipment starts and stops was calculated based on the mean and standard deviation of the temperature recovery time to quantify the stability and control effect of cold chain equipment operation. This impact coefficient comprehensively considers the degree of disturbance to the temperature environment caused by the frequent start-up and shutdown of cold chain equipment. If the impact coefficient is large, it means that the start-up and shutdown mode of the cold chain equipment may cause large temperature fluctuations, thereby accelerating the corrosion rate of agricultural products. If the impact coefficient is small, it indicates that the cold chain equipment is operating relatively smoothly and the temperature environment in the carriage is stable, which helps reduce the corrosion risk of agricultural products and provides a scientific basis for optimizing the start-up and shutdown strategy of cold chain equipment.

[0081] Step S4: construct an agricultural product corrosion rate impact assessment model based on the impact coefficient of frequent opening and closing of carriage doors and the impact coefficient of frequent start and stop of cold chain equipment, output an agricultural product corrosion rate impact assessment index, and assess the degree of increase in the impact of agricultural product corrosion rate, as follows:

[0082] An agricultural product corrosion rate impact assessment model is constructed based on the impact coefficient of frequent door opening and closing and the impact coefficient of frequent start-stop of cold chain equipment. The agricultural product corrosion rate impact assessment index (CPI) is output. The model is based on the following formula: CPI = a1*Cdoor+a2*Ccool, where a1 and a2 represent the preset proportional coefficients of the impact coefficient of frequent door opening and closing and the impact coefficient of frequent start-stop of cold chain equipment, respectively, and both a1 and a2 are greater than 0.

[0083] It should be noted that the above formulas are all dimensionless and calculated numerically. Common dimensionless methods include Min-Max normalization and Z-Score standardization, which are not described in detail here. a1 and a2 are set according to actual conditions. For example, the expert weighting method can be adopted, that is, experts in related fields are invited to conduct professional opinion surveys and comprehensive evaluations to determine the preset proportional coefficients of various indicators. For example, a1 and a2 can be 0.4 and 0.6 respectively.

[0084] From the above calculation expression, it can be seen that the greater the impact coefficient of frequent opening and closing of carriage doors and the impact coefficient of frequent start and stop of cold chain equipment, the greater the impact assessment index of agricultural product corrosion rate, indicating that the degree of increase in the impact of agricultural product corrosion rate is higher, that is, the negative impact of frequent opening and closing of carriage doors and frequent start and stop of cold chain equipment on the storage and transportation quality of agricultural products is more significant, which may lead to faster corrosion and higher loss rate. Conversely, the smaller the impact coefficient of frequent opening and closing of carriage doors and the impact coefficient of frequent start and stop of cold chain equipment, the smaller the impact assessment index of agricultural product corrosion rate, indicating that the carriage temperature environment is more stable, the operation of cold chain equipment is more stable, the agricultural products are less affected by temperature fluctuations, the corrosion risk is reduced, and the storage and transportation quality is improved;

[0085] The agricultural product corrosion rate impact assessment index is compared with the preset agricultural product corrosion rate impact assessment index threshold to assess the degree of increase in the impact of agricultural product corrosion rate, as follows:

[0086] If the agricultural product corrosion rate impact assessment index is greater than the agricultural product corrosion rate impact assessment index threshold, the current cold chain environment has a greater impact on the corrosion of agricultural products, the temperature stability of agricultural products is poor, and the corrosion risk is increasing. It is necessary to optimize the logistics distribution path and generate a high impact rising degree signal. If the agricultural product corrosion rate impact assessment index is less than or equal to the agricultural product corrosion rate impact assessment index threshold, it indicates that the current cold chain environment has a low impact on the corrosion of agricultural products, the temperature control is relatively stable, and the agricultural product corrosion rate is within an acceptable range. A low impact rising degree signal is generated and there is no need to optimize the logistics distribution path.

[0087] Step S5, improving the existing ant colony algorithm according to the degree of increase in the impact of the agricultural product corrosion rate to optimize the logistics distribution path of the agricultural products;

[0088] When a high-impact rising degree signal is generated, the existing ant colony algorithm is improved to optimize the logistics distribution path of agricultural products as follows:

[0089] Step B1, initialize the ant colony: set the number of ants ml, initialize pheromones: in represents the pheromone concentration of the path u→v, is a minimum positive value;

[0090] Step B2, path construction optimization: Each ant starts from the current delivery point and determines the probability of choosing the next stop based on the pheromone concentration and heuristic factor. The calculation formula is as follows: Where PL represents the probability of the ant choosing the next station, JH represents the set of next nodes that all current ants can choose. represents the pheromone concentration of the path u→v, η uv represents the heuristic factor of the path u→v, JL represents the actual distance from node u to node v, and CPI represents the agricultural product corrosion rate impact assessment index output by the current agricultural product corrosion rate impact assessment model;

[0091] Step B3, pheromone update: After each ant selects a path, the pheromone is updated immediately: Where ρ is the volatility constant, represents the pheromone increment, Where QS represents the pheromone constant, and GF represents the length of the path completed by the current ant;

[0092] It should be noted that the volatility constant is used to control the forgetting rate of pheromone, which is usually set to 0.1≤ρ≤0.5. The pheromone constant is usually set to: QS=10*Lbest, where Lbest represents the historical optimal path length;

[0093] Step B4: preset the maximum number of iterations, repeat steps B2 and B3 in a loop, and output the delivery path after the iteration stops;

[0094] It should be noted that the above formulas are all dimensionless and numerical calculations. Common dimensionless methods include Min-Max normalization and Z-Score normalization, which will not be described here.

[0095] The present invention constructs an initial distribution map with time window constraints based on the geographical coordinates of the distribution center and the delivery point, and monitors the distribution sequence status of agricultural products in real time during the distribution process. It can accurately identify the reversal of the distribution sequence caused by emergency situations. When the distribution sequence is in an abnormal state, the frequent opening and closing influence coefficient of the carriage door is calculated according to the acquired carriage door opening and closing information. At the same time, the frequent opening and closing influence coefficient of the cold chain equipment is calculated using the start and stop information of the cold chain equipment. Then, the agricultural product corrosion rate impact assessment model is constructed and the agricultural product corrosion rate impact assessment index is output, so as to quantitatively assess the degree of increase in the impact of the cold chain environment on the corrosion rate of agricultural products. Based on the agricultural product corrosion rate impact assessment index, the existing The ant colony algorithm is improved by comprehensively considering multiple objective factors such as path distance, frequent opening and closing of carriage doors, and start-up and shutdown of cold chain equipment in the path selection process, optimizing the pheromone update rules, and ensuring that the selected distribution path can minimize temperature fluctuations and equipment load while meeting the time window constraints, thereby reducing the risk of agricultural product corrosion, improving the stability of cold chain transportation and logistics distribution efficiency. It not only realizes real-time dynamic monitoring of the cold chain logistics environment, but also provides a precise control basis for the cold chain transportation of agricultural products through multi-objective optimization decision-making, significantly improving the storage quality and safety of agricultural products during transportation, while reducing energy consumption and equipment loss caused by unstable temperature control.

[0096] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0097] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0098] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A multi-objective agricultural product logistics distribution path optimization method based on ant colony algorithm, characterized by: The steps include: Step S1: constructing an initial distribution map with time window constraints based on the geographic coordinates of the distribution center and the delivery point, distributing the agricultural products according to the initial distribution map, and monitoring the distribution sequence status of the agricultural products, wherein the distribution sequence status of the agricultural products includes a normal distribution sequence status and an abnormal distribution sequence status; Step S2: When the delivery sequence of the agricultural products is in an abnormal delivery sequence, the opening and closing information of the carriage doors is obtained during the delivery of the agricultural products, and the frequent opening and closing influence coefficient of the carriage doors is calculated based on the opening and closing information of the carriage doors to evaluate the impact of the frequent opening and closing of the carriage doors on the corrosion rate of the agricultural products; Step S3: When the delivery order status of the agricultural products is in an abnormal delivery order state, the start and stop information of the cold chain equipment is obtained during the delivery process of the agricultural products, and the frequent start and stop impact coefficient of the cold chain equipment is calculated based on the start and stop information of the cold chain equipment to evaluate the impact of the frequent start and stop of the cold chain equipment on the corrosion rate of the agricultural products; Step S4: constructing an agricultural product corrosion rate impact assessment model based on the impact coefficient of frequent door opening and closing and the impact coefficient of frequent start and stop of cold chain equipment, outputting an agricultural product corrosion rate impact assessment index, and assessing the degree of increase in the impact of agricultural product corrosion rate; Step S5: improving the existing ant colony algorithm according to the degree of increase in the impact of the agricultural product corrosion rate to optimize the logistics distribution path of the agricultural products.

2. The multi-objective agricultural product logistics distribution path optimization method based on ant colony algorithm according to claim 1 is characterized by: The logic for obtaining the influence coefficient of frequent door opening and closing is as follows: During the agricultural product distribution process, the vehicle door opening and closing information is obtained, including the number of times the vehicle door is opened and closed, the duration of a single opening of the vehicle door, the temperature inside and outside the vehicle, and the opening area of the vehicle door. The heat exchange between the vehicle door and the outside world when the vehicle door is open is calculated as follows: Qj = dl*mj*ΔT*Topen, where Qj represents the heat exchange between the vehicle door and the outside world, dl represents the convection heat transfer coefficient, mj represents the opening area of the vehicle door, ΔT represents the temperature difference between the inside and outside of the vehicle, and Topen represents the duration of the vehicle door opening. Calculate the temperature fluctuation each time the door is opened: Where twi represents the temperature fluctuation when the door is opened, jl represents the air mass in the compartment, jl = md*tj, where md represents the air density and tj represents the interior volume of the compartment; The temperature fluctuation data set KB is constructed by the temperature fluctuation each time the door is opened. i }={twi1,twi2,...,twi I }, where twi i represents the temperature fluctuation when the i-th door is opened, i = {1, 2, ..., I}, I is the total number of times the door is opened; Calculate the correlation coefficient between the temperature fluctuation of the i-th compartment door opening and the temperature fluctuation of the remaining compartment door openings in the temperature fluctuation data set: Where xg represents the correlation coefficient between the temperature fluctuation of the i-th door opening and the temperature fluctuation of the j-th door opening in the temperature fluctuation data set, twi j represents the temperature fluctuation of the remaining j-th compartment door opening in the temperature fluctuation data set, and σt represents the standard deviation of the temperature fluctuation of all compartment door openings in the temperature fluctuation data set; Compare the correlation coefficient with a preset correlation coefficient threshold. If the correlation coefficient is greater than the correlation coefficient threshold, add the temperature fluctuation of the j-th door opening to the similar neighborhood set of the temperature fluctuation of the i-th door opening. Each door opening is defined as a node, and the correlation coefficient between two door opening and closing events is used as the edge weight to construct the door opening and closing-temperature fluctuation propagation network; Calculate the node influence as follows: Step A1: Initialize the initial node influence of all nodes: PR 0 (i) represents the node influence at iteration 0, i = {1, 2, ..., I}, I is the total number of door openings; Step A2: For each node i, calculate the new node influence in this round based on the node influence in the previous round: PR t+1 (i) represents the node influence at the t+1th iteration, N(j) represents the similar neighborhood set of node i, deg represents the total number of nodes in the similar neighborhood set, PR t (j) represents the node influence at the tth iteration, and d represents the damping factor; Step A3: Repeat step A2 to traverse all nodes and continuously update the node influence. The difference between the node influences of two consecutive rounds is compared with the preset convergence threshold. If the difference between the node influences of two consecutive rounds is less than the convergence threshold, the iteration is stopped. Calculate the influence coefficient of frequent opening and closing of carriage doors: Where Cdoor represents the influence coefficient of frequent door opening and closing, and PR(i) represents the node influence of the i-th door opening event.

3. The multi-objective agricultural product logistics distribution path optimization method based on ant colony algorithm according to claim 1 is characterized by: The logic for obtaining the impact coefficient of frequent start-up and shutdown of cold chain equipment is as follows: Obtain the compartment temperature each time the cold chain equipment is started, and compare the compartment temperature each time the cold chain equipment is started with the refrigeration temperature of the agricultural products. If the compartment temperature at the time the cold chain equipment is started is higher than the refrigeration temperature of the agricultural products, add the compartment temperature at the time of the cold chain equipment startup to the temperature data set; Calculate the temperature recovery time for the carriage temperature to return to the agricultural product refrigeration temperature in the temperature data set: τrec=min{t-ton|T(t)≤Tset,T(t)>Tset+δ}, where τrec represents the temperature recovery time, t-ton represents the time interval calculated from the cold chain equipment startup time ton, T(t) represents the carriage temperature at time t, Tset represents the refrigeration temperature of the agricultural product, and δ is the temperature offset value, indicating that the carriage temperature must exceed this value before cooling is triggered; Calculate the mean temperature recovery time for the carriage temperature to return to the agricultural product refrigeration temperature in the temperature data set: Where jz represents the mean temperature recovery time of the carriage temperature in the temperature data set to return to the agricultural product refrigeration temperature, τrec g represents the temperature recovery time for the g-th compartment temperature in the temperature data set to return to the agricultural product refrigeration temperature, g = {1, 2, ..., G}, where G is a positive integer; Calculate the standard deviation of the temperature recovery time for the carriage temperature to return to the agricultural product refrigeration temperature in the temperature data set: Where bz represents the standard deviation of the temperature recovery time for the compartment temperature to return to the agricultural product refrigeration temperature in the temperature data set; Calculate the impact coefficient of frequent start and stop of cold chain equipment: Ccool represents the impact coefficient of frequent start and stop of cold chain equipment.

4. The multi-objective agricultural product logistics distribution path optimization method based on ant colony algorithm according to claim 1 is characterized by: An agricultural product corrosion rate impact assessment model is constructed based on the impact coefficient of frequent opening and closing of carriage doors and the impact coefficient of frequent start and stop of cold chain equipment, and the agricultural product corrosion rate impact assessment index CPI is output. The model is based on the following formula: CPI = a1*Cdoor+a2*Ccool, where a1 and a2 represent the preset proportional coefficients of the impact coefficient of frequent opening and closing of carriage doors and the impact coefficient of frequent start and stop of cold chain equipment, respectively, and a1 and a2 are both greater than 0.

5. The multi-objective agricultural product logistics distribution path optimization method based on ant colony algorithm according to claim 4 is characterized by: The agricultural product corrosion rate impact assessment index is compared with the preset agricultural product corrosion rate impact assessment index threshold to assess the degree of increase in the impact of agricultural product corrosion rate, as follows: If the agricultural product corrosion rate impact assessment index is greater than the agricultural product corrosion rate impact assessment index threshold, a high impact rising degree signal is generated; if the agricultural product corrosion rate impact assessment index is less than or equal to the agricultural product corrosion rate impact assessment index threshold, a low impact rising degree signal is generated.

6. The multi-objective agricultural product logistics distribution path optimization method based on ant colony algorithm according to claim 5 is characterized by: When a high-impact rising degree signal is generated, the existing ant colony algorithm is improved to optimize the logistics distribution path of agricultural products as follows: Step B1, initializing the ant colony: setting the number of ants and initializing pheromones; Step B2, path construction optimization: Each ant starts from the current delivery point and determines the probability of choosing the next stop based on the pheromone concentration and heuristic factor; Step B3, pheromone update: after each ant selects a path, the pheromone is updated immediately; Step B4: preset the maximum number of iterations, repeat steps B2 and B3 in a loop, and output the delivery path after the iteration stops.

Images