Multi-index collaborative optimization method and device, medium and equipment

Abstract

The invention relates to a multi-index collaborative optimization method and device, a medium and equipment, and belongs to the field of logistics service quality. The method comprises the following steps: determining a main index, and determining each sub-index of the main index; obtaining logistics text data corresponding to each sub-index, and preprocessing the logistics text data; performing polynomial fitting according to the logistics text data corresponding to each sub-index, and determining an index polynomial for the main index; according to the index polynomial, constructing a multi-index collaborative optimization model corresponding to the main index; the multi-index collaborative optimization model is solved through a preset solving method, an optimal solution set for the main index is determined, and the optimal solution set is at least composed of partial sub-indexes in the sub-indexes; and determining a multi-index collaborative optimization strategy according to the optimization solution set, and executing the logistics service according to the multi-index collaborative optimization strategy, thereby improving the efficiency and quality of the logistics service.

Description

Technical Field

[0001] The present invention belongs to the field of logistics service quality and relates to a multi-index collaborative optimization method, device, medium and equipment. Background Art

[0002] As a core component of a company's supply chain, the efficiency and quality of logistics services directly impact a company's market competitiveness and customer satisfaction. Therefore, improving the efficiency and quality of logistics services is crucial.

[0003] At present, the efficiency and quality of logistics services are mainly affected by factors such as service quality, operational efficiency and customer satisfaction. Therefore, how to improve the efficiency and quality of logistics services while ensuring service quality, operational efficiency and customer satisfaction has become an increasingly important research focus.

[0004] In existing technologies, principal component analysis is commonly used to analyze the relationship between service quality, operational efficiency, customer satisfaction, and cost, as well as the efficiency and quality of logistics services. However, the analysis results obtained by this method are not ideal. Summary of the Invention

[0005] In view of this, the object of the present invention is to provide a multi-index collaborative optimization method and device to improve the efficiency and quality of logistics services.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] Option 1:

[0008] A multi-index collaborative optimization method specifically includes the following steps:

[0009] S1: Determine the main indicator and its sub-indicators;

[0010] S2: Obtain the logistics text data corresponding to each sub-indicator and preprocess it;

[0011] S3: Perform polynomial fitting based on the logistics text data corresponding to each sub-indicator to determine the indicator polynomial for the main indicator;

[0012] S4: Based on the indicator polynomial, a multi-indicator collaborative optimization model corresponding to the main indicator is constructed;

[0013] S5: Solve the multi-index collaborative optimization model using a preset solution method to determine an optimized solution set for the main index, where the optimized solution set is composed of at least some of the sub-indicators.

[0014] S6: Determine a multi-index collaborative optimization strategy based on the optimization solution set to perform logistics services according to the multi-index collaborative optimization strategy.

[0015] Furthermore, in step S1, parts of logistics service quality, customer satisfaction, operational efficiency, and cost control can be used as main indicators; factors affecting the main indicators can be used as sub-indicators of the main indicators; there is no limit on the number of main indicators. In the case of multiple main indicators, sub-indicators of each main indicator can be determined;

[0016] The sub-indicators of the main indicator F1, which is logistics service quality, can be logistics service quality complaint rate x1, cigarette damage rate x2, sorting error rate x3 and delivery damage rate x4; the sub-indicators of the main indicator F2, which is customer satisfaction, can be average complaint handling time x5, delivery staff to customer ratio x6 and high-quality staff ratio x7; the sub-indicators of the main indicator F3, which is operational efficiency, can be sorting equipment effective utilization rate x8, delivery vehicle per 10,000 boxes x9, and the ratio of special-shaped cigarettes x 10 , average delivery mileage per household x 11 and inventory turnover times x 12 The main indicator is cost control F4, and the sub-indicator can be single box warehouse cost x 13 , Single box sorting fee x 14 , Single box delivery fee x 15 And labor cost per box x 16 .

[0017] Furthermore, in step S2, the logistics text data corresponding to each sub-indicator is preprocessed, and the preprocessing includes identifying and processing abnormal text data and completing missing logistics text data.

[0018] Identify and address abnormal text data. Specifically, outliers may be caused by data entry errors, sensor failures, or other abnormal conditions. This paper uses statistical analysis methods, such as boxplots and Z-scores, to identify these outliers. For identified outliers, the paper chooses to delete, replace, or correct them based on their context and business logic.

[0019] To complete missing logistics text data, we use simple deletion for a small number of missing values. For a larger proportion of missing values, we use mean, median, or mode filling, as well as more advanced interpolation methods such as K-nearest neighbor (K-NN) or multiple interpolation methods.

[0020] Furthermore, in step S3, the polynomial fitting method is to use multiple linear regression fitting and back propagation (Bp) neural network fitting;

[0021] When the linear fitting of the indicator polynomial is determined, the indicator polynomial for the main indicator is re-determined based on the logistics text data corresponding to each sub-indicator until the indicator polynomial is no longer linearly fitted.

[0022] Furthermore, in step S4, for logistics service quality F1, customer satisfaction F2, operational efficiency F3, and cost control F4, their respective multi-index collaborative optimization models can be determined based on their respective index polynomials:

[0023]

[0024] Among them, c, e, d, and g are the coefficients of sub-indicators, n1, n2, n3, and n4 are the numbers of sub-indicators corresponding to the main indicators F1, F2, F3, and F4, respectively; α, β, χ, and δ are the constant terms obtained based on expert experience and historical data analysis;

[0025] The multi-indicator collaborative optimization model corresponding to the main indicator is constructed as: Max(F1,F2,F3,-F4).

[0026] Furthermore, in step S5, the multi-index collaborative optimization model is solved according to the non-dominated sorting genetic algorithm to determine each optimization solution set for the main index.

[0027] Furthermore, in step S6, when there are multiple main indicators, a multi-indicator collaborative optimization strategy is determined according to the weight ratio of each main indicator in the preset optimization solution set;

[0028] Determine the multi-index collaborative optimization strategy based on the optimization solution set, including:

[0029] For each optimized solution set, input the optimized solution set into the trained prediction model to obtain the predicted value of the main indicator output by the prediction model; determine whether the predicted value reaches the preset indicator threshold; if so, determine the multi-indicator optimization strategy based on the optimized solution set; otherwise, discard the optimized solution set;

[0030] Training the prediction model specifically includes: obtaining a training solution set and using the historical prediction values ​​of the training solution set as labels; inputting the training solution set into the prediction model to be trained to obtain the prediction value output by the prediction model to be trained; determining a loss value based on the difference between the prediction value output by the prediction model to be trained and the label, and training the prediction model to be trained with the minimum loss value as the optimization goal, wherein the difference is positively correlated with the loss value.

[0031] Option 2:

[0032] A multi-index collaborative optimization device, comprising:

[0033] A determination module is used to determine the main indicator and the sub-indicators of the main indicator;

[0034] The acquisition module is used to obtain the logistics text data corresponding to each sub-indicator;

[0035] The fitting module is used to perform polynomial fitting based on the logistics text data corresponding to each sub-indicator to determine the indicator polynomial for the main indicator;

[0036] A construction module is used to construct a multi-indicator collaborative optimization model corresponding to the main indicator based on the indicator polynomial;

[0037] A solution module is used to solve the multi-index collaborative optimization model by a preset solution method, and determine an optimized solution set for the main index, where the optimized solution set is composed of at least some of the sub-indicators;

[0038] A formulation module is used to determine a multi-index collaborative optimization strategy based on an optimization solution set, so as to perform logistics services according to the multi-index optimization strategy.

[0039] Option 3:

[0040] A computer-readable storage medium stores a computer program, which implements the method of solution 1 when executed by a processor.

[0041] Option 4:

[0042] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the method of solution 1 is implemented when the processor executes the program.

[0043] The beneficial effects of this invention lie in: by performing polynomial fitting on the data of each sub-indicator corresponding to the main indicator, the present invention determines the indicator polynomial, which can then demonstrate the relationship between the sub-indicators and the main indicator. Furthermore, the indicator polynomial summarizes a multi-indicator collaborative optimization model, which can determine which sub-indicators have the greatest impact on the main indicator and reveal the patterns and relationships between different sub-indicators. Based on the determined optimal solution set, a collaborative optimization strategy is then formulated to optimize and improve the efficiency and quality of logistics services.

[0044] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0046] Figure 1 A flowchart of a multi-index collaborative optimization method provided by an embodiment of the present invention;

[0047] Figure 2 A fitting graph of a multiple linear regression fitting provided by an embodiment of the present invention;

[0048] Figure 3 A heat map provided by an embodiment of the present invention;

[0049] Figure 4 A schematic diagram of a multi-index collaborative optimization device provided by an embodiment of the present invention;

[0050] Figure 5 The embodiment of the present invention provides a method corresponding to Figure 1 Schematic diagram of the structure of the electronic equipment. DETAILED DESCRIPTION

[0051] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0052] See also Figures 1 to 5 , Figure 1 A flowchart of a multi-index collaborative optimization method provided by an embodiment of the present invention, the method specifically includes the following steps:

[0053] S1: Determine the main indicator and the sub-indicators of the main indicator.

[0054] In this embodiment, the process of implementing multi-index collaborative optimization in logistics services can be performed by a server, which performs the multi-index optimization and determines the multi-index collaborative optimization strategy. Of course, this specification does not limit the device that performs this process. For ease of description, the following description uses the server as the execution entity.

[0055] In this embodiment, when considering logistics services, it is usually analyzed from the aspects of logistics service quality, customer satisfaction, operational efficiency and cost control. Therefore, in the subsequent steps of this embodiment, the analysis of logistics service quality, customer satisfaction, operational efficiency and cost control is used as an example for explanation.

[0056] Logistics service quality refers to the performance of logistics services within the supply chain, including metrics such as the logistics service quality complaint rate, item damage rate, and sorting error rate. Customer satisfaction reflects the overall customer satisfaction with logistics services, including metrics such as the average complaint handling time and the percentage of satisfied customers. Operational efficiency, which measures the balance between resource utilization and transportation cost control, encompasses metrics such as the effective utilization rate of sorting equipment and the efficiency of delivery vehicle allocation. It provides a comprehensive assessment of the efficiency and economic viability of logistics operations. Cost control refers to various costs involved in the logistics service process, such as inventory and labor costs.

[0057] The server can identify logistics service quality, customer satisfaction, operational efficiency, and cost control as primary indicators. For example, logistics service quality can be identified as the primary indicator. After determining the primary indicator, factors influencing the primary indicator can be analyzed to determine the primary indicator. Further analysis of logistics services can be conducted based on the primary indicator. Factors influencing the primary indicator can be used as sub-indicators of the primary indicator. For example, if logistics service quality is the primary indicator, the logistics service complaint rate, item damage rate, and sorting error rate can all be sub-indicators.

[0058] Of course, the number of main indicators is not limited in this description. When there are multiple main indicators, sub-indicators of the main indicator are determined for each main indicator.

[0059] In this embodiment, the sub-indicators of the main indicator F1 of logistics service quality can be logistics service quality complaint rate x1, cigarette damage rate x2, sorting error rate x3 and delivery damage rate x4; the sub-indicators of the main indicator F2 of customer satisfaction can be average complaint handling time x5, delivery personnel to customer ratio x6 and high-quality personnel ratio x7; the sub-indicators of the main indicator F3 of operational efficiency can be sorting equipment effective utilization rate x8, delivery vehicle per 10,000 boxes x9, and the ratio of special-shaped cigarettes x 10 , average delivery mileage per household x 11 and inventory turnover times x 12 The main indicator is cost control F4, and the sub-indicator can be single box warehouse cost x 13 , Single box sorting fee x 14 , Single box delivery fee x 15 And labor cost per box x 16 .

[0060] S2: Obtain the logistics text data corresponding to each sub-indicator.

[0061] In this embodiment, after determining the main indicator and each sub-indicator of the main indicator, the server obtains the logistics text data corresponding to each sub-indicator to further analyze the main indicator.

[0062] Of course, the server may also pre-process the logistics text data corresponding to each sub-indicator, and the pre-processing may include at least one of processing abnormal text data, completing missing text data, and data conversion.

[0063] S3: Perform polynomial fitting based on the logistics text data corresponding to each sub-indicator to determine the indicator polynomial for the main indicator.

[0064] In this embodiment, the server performs polynomial fitting based on the logistics text data corresponding to each sub-indicator to determine the indicator polynomial for the main indicator.

[0065] For example, if the primary indicator is logistics service quality F1, and the sub-indicators are logistics service quality complaint rate X1, cigarette breakage rate X2, sorting error rate X3, and delivery damage rate X4, and historical data suggests that logistics service quality and logistics service quality complaint rate X1 are identical, then the corresponding logistics text data for each sub-indicator can be used to fit the equations representing F1 and X2, X3, and X4, thus forming the indicator polynomial for F1.

[0066] In this embodiment, the polynomial fitting method may be multiple linear regression fitting and back propagation (BP) neural network fitting.

[0067] Taking multiple linear regression fitting as an example, Figure 2 This is a fitted graph of a multiple linear regression fit provided in this embodiment. The dark line is the predicted value line of the multiple linear regression fit, the light line is the actual value of the indicator, the vertical axis is the numerical representation of logistics service quality, and the horizontal axis is the indicator value for 30 months.

[0068] The following is an example of an indicator polynomial fitted by multiple linear regression, such as logistics service quality:

[0069] F1=0.018x 11 -3.4x2-2.2x3-2.5x4-6.6e -4 x5-25.0x6-0.4x7-0.11x8-0.012x9-1.6e - 3 x 10 -100.0x1-0.2x 12 +1.2e -3 x 13 +3.1e -3 x 14 -1.5e -3 x 15 -3.9e -4 x 16 +1.9

[0070] It's worth noting that when fitting data using a BP neural network, polynomials can be determined for both linear and nonlinear fits. While linear models struggle to handle complex nonlinear relationships, nonlinear fitting can capture and express these complex nonlinear relationships between variables. It can fit curves or non-linear relationships that linear models can't capture. Nonlinear models offer greater flexibility and can be adjusted to various forms, such as polynomials, exponentials, and logarithms, to accommodate changing data patterns.

[0071] Although nonlinear fitting has obvious advantages in fitting capabilities, it is also computationally expensive and requires more resources and time. Moreover, the parameter estimation and optimization process of nonlinear models is more complicated and may require the use of more complex numerical optimization methods.

[0072] Therefore, in this embodiment, considering the complexity and variability of the data and the need for intuitive display of the results, both linear and nonlinear fitting methods are used. This not only improves the model's fit to the data but also ensures that the results are easy to understand and display, thus achieving a balance between accuracy and interpretability. In this way, the advantages of both fitting methods can be fully utilized to obtain more reliable and accurate prediction results.

[0073] In this embodiment, when the server determines that the index polynomial is linearly fitted, it re-determines the index polynomial for the main index based on the logistics text data corresponding to each sub-indicator until the index polynomial is no longer linearly fitted. For example, the Durbin-Watson value of the index polynomial is calculated to determine whether it is linearly fitted.

[0074] S4: Based on the indicator polynomial, construct a multi-indicator optimization model corresponding to the main indicator.

[0075] In this embodiment, the server constructs a multi-index collaborative optimization model corresponding to the main index based on the fitted index polynomial.

[0076] In this embodiment, for logistics service quality F1, customer satisfaction F2, operational efficiency F3, and cost control F4, respective multi-index collaborative optimization models can be determined based on their respective index polynomials, as shown below:

[0077]

[0078] Among them, c, e, d, g are the coefficients of sub-indicators. 16 The constraints are:

[0079] Upper bound: lb = [0; 0; 0; 0; 0.004; 0.1; 0.9; 6; 15; 1; 1; 20; 20; 20; 50;

[0080] Lower bound: ub = [0.006; 0.2; 0.1; 0.15; 90; 0.01; 0.6; 1; 40; 45; 5; 2.5; 150; 150; 300; 500].

[0081] In this embodiment, when logistics service is determined as the main indicator, the sub-indicators affecting logistics service are logistics service quality F1, customer satisfaction F2, operational efficiency F3, and cost control F4. At this time, a multi-indicator collaborative optimization model for logistics service can be determined as follows:

[0082] Max(F1,F2,F3,-F4)

[0083] S5: Solve the multi-index collaborative optimization model through a preset solution method to determine an optimization solution set for the main index, where the optimization solution set is composed of at least some of the sub-indicators.

[0084] In this embodiment, the server solves the multi-indicator collaborative optimization model of the main indicator through a preset solution method, and determines the optimization solution set for the main indicator. The optimization solution set is composed of at least some of the sub-indicators in each sub-indicator. That is, by solving the above-mentioned multi-indicator collaborative optimization model, the optimization solution set of the multi-indicator collaborative optimization model corresponding to the main indicator can be solved.

[0085] In this embodiment, the server may solve the multi-index collaborative optimization model according to a non-dominated sorting genetic algorithm to determine each optimization solution set for the main index.

[0086] S6: Determine a multi-index collaborative optimization strategy based on the optimization solution set to execute logistics services according to the multi-index collaborative optimization strategy.

[0087] When there are multiple main indicators, the multi-indicator collaborative optimization strategy is determined according to the weight ratio of each main indicator in the preset optimization solution.

[0088] Furthermore, for each optimized solution set, the server inputs the optimized solution set into the trained prediction model (i.e., BP neural network) to obtain the predicted value of the main indicator output by the prediction model. It determines whether the predicted value reaches the preset indicator threshold. If so, the multi-indicator collaborative optimization strategy is determined based on the optimized solution set. Otherwise, the optimized solution set is discarded. Specifically, the prediction model first receives pre-processed logistics text data as input, then extracts features through multiple hidden layers, and finally generates the predicted value of the main indicator in the output layer. Through a carefully designed network structure, the BP neural network can capture the nonlinear relationships in the data, thereby providing more accurate prediction results.

[0089] Among them, the following method is used to train the prediction model:

[0090] Obtain a training solution set and use its historical prediction values ​​as labels. Input the training solution set into the prediction model to be trained to obtain the predicted values ​​output by the prediction model. Determine the loss value based on the difference between the predicted values ​​output by the prediction model to be trained and the labels. Train the prediction model to be trained with the minimum loss value as the optimization goal. The difference is positively correlated with the loss value.

[0091] In this specification, the method for determining the loss value by calculating the difference between the predicted value and the annotation is not limited. For example, the mean square error (MSE) loss is calculated as follows:

[0092]

[0093] Among them, n is the number of main indicators (number of optimized solution sets), y i is the true observation value (label) of the i-th main indicator, a i is the model's predicted value for the i-th main indicator.

[0094] It is worth noting that in order to reduce the impact of outliers in the optimized solution set, the mean absolute error (MAE) loss with more stable errors can also be used. The calculation method is as follows:

[0095]

[0096] Among them, n is the number of main indicators (number of optimized solution sets), y i is the true observation value (label) of the i-th main indicator, a i is the model's predicted value for the i-th main indicator.

[0097] In this embodiment, the server determines a multi-index collaborative optimization strategy based on the determined optimization solution set, so as to perform logistics services according to the multi-index collaborative optimization strategy.

[0098] In this embodiment, after determining the four indicators (including main indicators or sub-indicators) that need to be considered for the evaluation object, a weight, that is, a subjective preference, is assigned to each indicator based on its importance in the strategy. The weight is usually a value between 0 and 1, and the sum of all indicator weights should be equal to 1. Secondly, each strategy is scored on each indicator. The score is usually a quantitative value, which can be a score from 1 to 10, or other values ​​determined based on the characteristics of the indicator. Thirdly, when calculating the weighted score, the score of each plan on each indicator is multiplied by its corresponding weight, and then these products are added together to obtain the total weighted score of each strategy.

[0099] For example, for the four main indicators: logistics service quality, customer satisfaction, operational efficiency, and cost control, each main indicator is assigned a corresponding weight based on its importance in the overall decision-making. Specifically, logistics service quality has the highest weight, 0.7, indicating its importance in the strategy; customer satisfaction and operational efficiency each have a weight of 0.1, and cost control also has a weight of 0.1.

[0100] In this embodiment, the method for determining the weight ratio of each main indicator in the preset optimization solution set may be an entropy weight method.

[0101] In this embodiment, high logistics service quality, high customer satisfaction, or high per capita delivery volume and low single-box logistics cost are selected as goals respectively. In the determined optimization solution set, the logistics service quality priority scenario, customer satisfaction priority scenario, operational efficiency priority scenario and cost control priority scenario can be set to screen out the optimization solution set that is more suitable for the goal and determine a better multi-indicator collaborative optimization strategy.

[0102] In a scenario where logistics service quality is prioritized, a multi-indicator collaborative optimization model for logistics service quality can be optimized by reducing coefficients such as the complaint rate, cigarette breakage rate, sorting error rate, delivery damage rate, delivery personnel to customer ratio, and inventory turnover rate to improve logistics service quality. Coefficients such as the effective utilization rate of sorting equipment, the proportion of cigarettes with unusual shapes, and single-case sorting can also be increased to improve logistics service quality.

[0103] In a scenario where customer satisfaction is prioritized, a multi-indicator collaborative optimization model for customer satisfaction can be optimized by reducing factors such as the complaint rate, cigarette damage rate, and the delivery staff-to-customer ratio to improve customer satisfaction. Factors such as the sorting error rate, delivery damage rate, the proportion of highly qualified personnel, and the effective utilization rate of sorting equipment can also be increased to further improve customer satisfaction.

[0104] In a scenario where operational efficiency is prioritized, the multi-indicator collaborative optimization model prioritizing operational efficiency can be optimized by reducing coefficients such as the complaint rate, cigarette breakage rate, vehicle allocation per 10,000 cases, single-case sorting, single-case delivery, and single-case labor to improve operational efficiency. It can also be optimized by increasing coefficients such as the sorting error rate, delivery damage rate, average complaint handling time, delivery personnel to customer ratio, proportion of highly qualified personnel, effective utilization rate of sorting equipment, proportion of cigarettes with unusual shapes, and average delivery mileage per household to improve operational efficiency.

[0105] In a cost-prioritized multi-indicator collaborative optimization model, we can optimize cost control by reducing factors such as the complaint rate, cigarette breakage rate, inventory turnover, and single-case storage coefficients. We can also increase factors such as the sorting error rate, delivery damage rate, delivery personnel to customer ratio, the proportion of highly qualified personnel, the effective utilization rate of sorting equipment, the number of vehicles allocated for 10,000-case deliveries, single-case delivery, and single-case labor costs to improve cost control.

[0106] In this embodiment, an optimization solution set table is exemplarily given, as shown in Table 1, wherein solutions of type 1 for logistics service quality, type 2 for customer satisfaction, type 3 for operational efficiency, and type 4 for cost control are preferred.

[0107] Table 1 Optimized solution set table

[0108]

[0109]

[0110] Based on the solution in Table 1, the linear correlation between logistics service quality, customer satisfaction, operational efficiency, and cost control can be calculated to obtain the correlation matrix shown in Table 2. The data can be visualized and plotted into a thermal Figure 3 .

[0111] Table 2 Correlation matrix

[0112] Logistics service quality Customer satisfaction operational efficiency Cost Control Logistics service quality 1 0.2133 -0.8491 0.8842 Customer satisfaction 0.2133 1 -0.536 0.2519 operational efficiency -0.8491 -0.536 1 -0.9487 Cost control rate 0.8842 0.2519 -0.9487 1

[0113] According to thermal Figure 3 , we can see a strong positive correlation between logistics service quality and cost control. A strong negative correlation between logistics service quality and operational efficiency, and between operational efficiency and cost control. A moderate negative correlation between customer satisfaction and operational efficiency. A weak positive correlation between customer satisfaction and cost control.

[0114] Interpreting the correlation coefficients between the aforementioned variables: The correlation between logistics service quality and operational efficiency is -0.8491, indicating a strong negative correlation between the two variables. Specifically, as logistics service quality increases, operational efficiency tends to decrease. The correlation between logistics service quality and cost control is 0.8842, indicating a strong positive correlation between the two variables. Specifically, as logistics service quality increases, cost control tends to increase. The correlation between customer satisfaction and operational efficiency is -0.536, indicating a moderate negative correlation. Specifically, as customer satisfaction increases, operational efficiency may decrease, suggesting a certain inverse relationship between the two. The correlation between customer satisfaction and cost control is 0.2519, indicating a weak positive correlation between the two variables. This suggests that increasing customer satisfaction does not necessarily significantly affect cost control. The correlation between operational efficiency and cost control is -0.948669, indicating a very strong negative correlation. Specifically, as operational efficiency increases, cost control almost always decreases, indicating that the two variables have completely opposite trends.

[0115] The above is a multi-index collaborative optimization method provided by one or more embodiments of the present invention. Based on the same idea, the present invention also provides a corresponding multi-index collaborative optimization device, such as Figure 4 shown.

[0116] Figure 4 This is a schematic diagram of a multi-index collaborative optimization device provided in this embodiment, which specifically includes:

[0117] A determination module is used to determine the main indicator and the sub-indicators of the main indicator;

[0118] The acquisition module is used to obtain the logistics text data corresponding to each sub-indicator;

[0119] A fitting module, configured to perform polynomial fitting based on the logistics text data corresponding to each sub-indicator to determine an indicator polynomial for the main indicator;

[0120] A construction module is used to construct a multi-indicator collaborative optimization model corresponding to the main indicator based on the indicator polynomial;

[0121] A solution module, configured to solve the multi-index collaborative optimization model using a preset solution method, and determine an optimized solution set for the main index, wherein the optimized solution set is composed of at least some of the sub-indicators;

[0122] A formulation module is used to determine a multi-index collaborative optimization strategy according to an optimization solution set, so as to perform logistics services according to the multi-index collaborative optimization strategy.

[0123] The solution module is used to solve the multi-index collaborative optimization model according to the non-dominated sorting genetic algorithm, and determine each optimization solution set for the main index.

[0124] A formulation module is used to input each optimized solution set into the trained prediction model, obtain the predicted value of the main indicator output by the prediction model, and judge whether the predicted value reaches the preset indicator threshold. If so, a multi-indicator collaborative optimization strategy is determined based on the optimized solution set; if not, the optimized solution set is discarded.

[0125] Optionally, the device also includes a training module for obtaining a training solution set, and using the historical prediction values ​​of the training solution set as labels, inputting the training solution set into the prediction model to be trained, obtaining the prediction value output by the prediction model to be trained, determining the loss value based on the difference between the prediction value output by the prediction model to be trained and the label, and training the prediction model to be trained with the minimum loss value as the optimization goal, and the difference is positively correlated with the loss value.

[0126] Specifically, the determination module is used to determine the multi-indicator collaborative optimization strategy according to the weight ratio of each main indicator in the preset optimization solution set when there are multiple main indicators.

[0127] Specifically, the fitting module is used to preprocess the logistics text data corresponding to each sub-indicator, and the preprocessing includes at least one of processing abnormal text data, completing missing text data, and data conversion.

[0128] Specifically, the construction module is used to, when determining the linear fitting of the indicator polynomial, re-determine the indicator polynomial for the main indicator according to the logistics text data corresponding to each sub-indicator, until the indicator polynomial is not linearly fitted.

[0129] The present invention also provides a computer-readable storage medium, which stores a computer program, which can be used to execute the above Figure 1 A multi-index collaborative optimization method is provided.

[0130] The present invention also provides a Figure 5 The electronic device shown has a structure such as Figure 5 As shown, at the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to achieve the above Figure 1 The multi-index collaborative optimization method.

[0131] Of course, in addition to software implementation, the present invention does not exclude other implementation methods, such as logic devices or a combination of software and hardware, etc., that is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0132] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. A multi-index collaborative optimization method, characterized in that: The specific steps include: S1: Determine the main indicator and its sub-indicators; S2: Obtain the logistics text data corresponding to each sub-indicator and pre-process it; S3: Perform polynomial fitting based on the logistics text data corresponding to each sub-indicator to determine the indicator polynomial for the main indicator; S4: Based on the indicator polynomial, a multi-indicator collaborative optimization model corresponding to the main indicator is constructed; S5: solving the multi-index collaborative optimization model by a preset solution method, and determining an optimization solution set for the main index, wherein the optimization solution set is composed of at least some of the sub-indicators; S6: Determine a multi-index collaborative optimization strategy according to the optimization solution set to execute logistics services according to the multi-index collaborative optimization strategy.

2. The multi-index collaborative optimization method according to claim 1, characterized in that: In step S1, some of logistics service quality, customer satisfaction, operational efficiency and cost control can be used as main indicators; factors affecting the main indicators are used as sub-indicators of the main indicators; there is no limit on the number of main indicators. When there are multiple main indicators, for each main indicator, the sub-indicators of the main indicator are determined; The main indicator is logistics service quality F1, and its sub-indicators are logistics service quality complaint rate x1, cigarette damage rate x2, sorting error rate x3 and delivery damage rate x4; the main indicator is customer satisfaction F2, and its sub-indicators are average complaint processing time x5, distribution staff to customer ratio x6 and high-quality staff ratio x7; the main indicator is operational efficiency F3, and its sub-indicators are sorting equipment effective utilization rate x8, delivery vehicle per 10,000 boxes x9, and special-shaped cigarette ratio x1. 10 , average delivery mileage per household x 11 and inventory turnover times x 12 ; The main indicator is cost control F4, and the sub-indicator is single box warehouse cost x 13 , Single box sorting fee x 14 , Single box delivery fee x 15 And labor cost per box x 16 .

3. The multi-index collaborative optimization method according to claim 1, characterized in that: In step S2, the logistics text data corresponding to each sub-indicator is preprocessed, and the preprocessing includes identifying and processing abnormal text data and completing missing logistics text data.

4. The multi-index collaborative optimization method according to claim 1, characterized in that: In step S3, the polynomial fitting method is to use multiple linear regression fitting and back propagation neural network fitting; When the linear fitting of the indicator polynomial is determined, the indicator polynomial for the main indicator is re-determined according to the logistics text data corresponding to each sub-indicator until the indicator polynomial is not linearly fitted.

5. The multi-index collaborative optimization method according to claim 2, characterized in that: In step S4, for logistics service quality F1, customer satisfaction F2, operational efficiency F3, and cost control F4, their respective multi-index collaborative optimization models can be determined according to their respective index polynomials: Among them, c, e, d, and g are the coefficients of sub-indicators, n1, n2, n3, and n4 are the numbers of sub-indicators corresponding to the main indicators F1, F2, F3, and F4, and α, β, χ, and δ are the constant terms obtained based on expert experience and historical data analysis; The multi-indicator collaborative optimization model corresponding to the main indicator is constructed as: Max(F1,F2,F3,-F4).

6. The multi-index collaborative optimization method according to claim 1, characterized in that: In step S5, the multi-index collaborative optimization model is solved according to the non-dominated sorting genetic algorithm to determine each optimization solution set for the main index.

7. The multi-index collaborative optimization method according to claim 1, characterized in that: In step S6, when there are multiple main indicators, a multi-indicator collaborative optimization strategy is determined according to the weight ratio of each main indicator in the preset optimization solution set; Determine the multi-index collaborative optimization strategy based on the optimization solution set, including: For each optimized solution set, the optimized solution set is input into the trained prediction model to obtain the predicted value of the main indicator output by the prediction model; whether the predicted value reaches the preset indicator threshold is determined; if so, the multi-indicator optimization strategy is determined based on the optimized solution set; otherwise, the optimized solution set is discarded; Training the prediction model specifically includes: obtaining a training solution set, and using the historical prediction values ​​of the training solution set as labels; inputting the training solution set into the prediction model to be trained to obtain the prediction value output by the prediction model to be trained; determining the loss value according to the difference between the prediction value output by the prediction model to be trained and the label, and training the prediction model to be trained with the minimum loss value as the optimization goal, wherein the difference is positively correlated with the loss value.

8. A multi-index collaborative optimization device, characterized in that: include: A determination module is used to determine the main indicator and the sub-indicators of the main indicator; The acquisition module is used to obtain the logistics text data corresponding to each sub-indicator; A fitting module is used to perform polynomial fitting based on the logistics text data corresponding to each sub-indicator to determine the indicator polynomial for the main indicator; A construction module is used to construct a multi-index collaborative optimization model corresponding to the main index according to the index polynomial; A solution module is used to solve the multi-index collaborative optimization model by a preset solution method, and determine an optimization solution set for the main index, where the optimization solution set is composed of at least some of the sub-indicators; A formulation module is used to determine a multi-index collaborative optimization strategy based on an optimization solution set, so as to perform logistics services according to the multi-index optimization strategy.

9. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

Images