Method for site selection and path planning of multiple distribution logistics center sites

Abstract

The invention relates to the technical field of digital data processing, in particular to a multi-distribution logistics center site selection and path planning method, which comprises the following steps: calculating the position of a logistics center based on a distribution distance and historical express quantity data of a distribution terminal node; calculating the basic logistics quantity of the logistics center based on the fluctuation condition of the total distribution quantity in one year; predicting the express quantity data of the current year based on the historical express quantity data by using an autoregression integral moving average model, and then calculating overflow logistics quantity data based on the express quantity data of the current year and the basic logistics quantity; acquiring data of rentable distribution logistics stations in a preset range near the logistics center position; matching a rented logistics site combination with the lowest cost based on the overflow logistics quantity data; and dynamically allocating corresponding distribution terminal nodes based on the rented logistics stations. Therefore, the flexibility and response speed of site selection and path planning of the logistics center are improved, and the operation cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of digital data processing, and in particular to a method for site selection and route planning of multiple distribution logistics centers. Background Art

[0002] Multi-distribution logistics center site selection refers to the process of selecting the best geographical locations for multiple logistics centers when planning a logistics network, aiming to optimize distribution efficiency, reduce costs and improve customer service. This process takes into account a variety of factors, including but not limited to transportation costs, facility costs, customer demand distribution, traffic conditions, land availability and cost, and the possibility of future expansion.

[0003] The existing logistics network is generally fixed, but the existing express delivery volume fluctuates greatly with the platform activities. As a result, there is a tendency for idle inventory to exist during the low period of online shopping, and insufficient inventory during the peak period, thereby reducing management efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide a method for site selection and route planning of multiple distribution logistics centers, aiming to improve the flexibility and response speed of site selection and route planning of logistics centers, thereby reducing operating costs.

[0005] To achieve the above-mentioned object, the present invention provides a method for site selection and route planning of multiple distribution logistics centers, including obtaining historical express delivery volume data of each distribution terminal node in the distribution area;

[0006] Calculate the location of the logistics center based on the delivery distance and historical express volume data of the delivery terminal nodes;

[0007] Calculate the basic logistics volume of the logistics center based on the fluctuation of the total distribution volume within a year;

[0008] Based on the historical express volume data, the autoregressive integrated moving average model is used to predict the current year's express volume data, and then the overflow logistics volume data is calculated based on the current year's express volume data and the basic logistics volume;

[0009] Obtaining data on leasable distribution logistics sites within a preset range near the logistics center location, the logistics site data including area price and facility type;

[0010] Match the combination of rental logistics sites with the lowest cost based on overflow flow data;

[0011] Dynamically allocate corresponding distribution terminal nodes based on rented logistics sites.

[0012] The specific steps of obtaining the historical express delivery volume data of each delivery terminal node in the delivery area include:

[0013] Collect logistics data from different channels of each distribution terminal node;

[0014] Clean up duplicates, outliers, or incomplete records in logistics data;

[0015] Format the cleaned logistics data to obtain historical express delivery volume data.

[0016] The specific steps of calculating the location of the logistics center based on the delivery distance and the historical express volume data of the delivery terminal node include:

[0017] Obtain all potential logistics center locations and obtain the data of the candidate center group;

[0018] Calculate the transportation time from each distribution terminal node to each candidate center in the candidate center group data;

[0019] The location of logistics centers is calculated using the centroid method based on transportation time and historical express volume data.

[0020] The specific steps of calculating the basic logistics volume of the logistics center based on the fluctuation of the total distribution volume within a year include:

[0021] Obtain the delivery data of all delivery terminal nodes in the previous year based on historical express delivery volume data;

[0022] Get holiday sales data based on delivery data;

[0023] Get the median of holiday express delivery volume data to replace all holiday express delivery volume data;

[0024] The basic logistics volume is calculated for each selected time period throughout the year based on the replaced holiday express volume data.

[0025] The specific steps of predicting the current year's express volume data based on the historical express volume data using the autoregressive integrated moving average model, and then calculating the overflow flow data based on the current year's express volume data and the basic logistics volume include:

[0026] Use the ADF test to determine whether the time series needs to be differentiated to make it stationary;

[0027] Use autocorrelation and partial autocorrelation plots to estimate p and q values ​​to build an ARIMA model;

[0028] Train the ARIMA model using historical express delivery volume data to ensure that the model can capture the trend, seasonality, and cyclical components in the data;

[0029] Evaluate the prediction accuracy of the model through cross-validation;

[0030] Use the trained ARIMA model to predict the express delivery volume of the current year on a monthly basis to obtain the predicted logistics volume;

[0031] When the predicted logistics volume exceeds the basic logistics volume, the excess is the overflow logistics volume.

[0032] The specific steps of obtaining the data of the leasable distribution logistics sites within a preset range near the logistics center location include:

[0033] Set the search radius based on the location of the logistics center;

[0034] Map the logistics center location and search radius to the logistics real estate database for search;

[0035] Based on the information obtained from the search, all eligible candidate sites are listed to obtain the data of rentable distribution logistics sites.

[0036] The specific steps of matching the combination of leased logistics sites with the lowest cost based on the overflow flow data include:

[0037] Obtain spill flow data;

[0038] Calculate the amount and type of additional storage required at various time periods based on spill flow data;

[0039] Based on the additional storage volume and type, matching is performed in the data of available distribution logistics sites to obtain the logistics site plan for selection;

[0040] Calculate the cost data of each candidate logistics station plan and arrange them in order from low to high cost.

[0041] The specific steps of calculating the cost data of each candidate logistics station solution and arranging them in order from low to high cost include:

[0042] For each candidate logistics site, collect various cost data;

[0043] Convert different types of costs in cost data into uniform units;

[0044] Calculate fixed costs and variable costs based on the converted data to obtain the logistics costs to be selected;

[0045] Arrange the logistics costs to be selected from low to high.

[0046] The specific steps of dynamically allocating corresponding distribution terminal nodes based on the leased logistics sites include:

[0047] Calculate the actual road distance from each distribution terminal node to each logistics site;

[0048] Choose the nearest logistics station for delivery.

[0049] A method for site selection and path planning of a multi-distribution logistics center of the present invention includes collecting and analyzing the express delivery volume information of all distribution terminal nodes in the distribution area in the past period of time. These data are the basis for evaluating the performance of the existing logistics network and the key basis for predicting future trends. By counting the express delivery volume of each node, high-density distribution areas and service weaknesses can be identified, providing a reference for the subsequent site selection of the logistics center. Next, based on the acquired historical express delivery volume data and the distance between each distribution terminal node, a mathematical algorithm or model (such as the center of gravity method, simulated annealing algorithm, etc.) is used to determine the optimal logistics center location. The ideal logistics center should be located in a position that can minimize the total distribution distance or the lowest distribution cost, thereby improving distribution efficiency and reducing transportation costs. In order to more accurately plan the scale and service capacity of the logistics center, it is also necessary to consider the changing trend of the total distribution volume within a year. This step involves calculating the fluctuation of the express delivery volume throughout the year to obtain the basic logistics volume that the logistics center needs to handle in different seasons or months. Such considerations help avoid the situation where the logistics center is overloaded during peak periods or idle resources during low periods. The autoregressive integrated moving average model (ARIMA) or other time series forecasting techniques are used to predict the express delivery volume for the next year based on historical express delivery volume data. The forecast results will be used to guide the operational planning of the logistics center, especially to make advance preparations for possible express delivery volume peaks. On this basis, the part exceeding the basic logistics volume, the so-called overflow volume, is calculated. This part of the logistics volume usually requires additional logistics sites to assist in processing to ensure that the service quality is not affected. The system will collect information on all logistics sites that can be rented within a certain range around the logistics center, including but not limited to area, rental price, facility type, etc. The purpose is to find a logistics site combination that can meet the overflow flow processing needs and maintain cost-effectiveness optimization. This process may involve complex multi-objective optimization problems, which require comprehensive consideration of multiple factors such as cost, location, and facility adaptability. Once the best logistics site combination is determined, the distribution tasks can be dynamically allocated to the corresponding distribution terminal nodes. This dynamic allocation mechanism can be flexibly adjusted according to the real-time changes in express delivery volume, ensuring that each logistics site can operate efficiently within its processing capacity while minimizing the operating cost of the entire distribution system.

[0050] In summary, this method not only improves the scientificity and rationality of logistics center site selection and route planning, but also enhances the flexibility and responsiveness of the logistics system, thereby reducing operating costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0052] Figure 1 It is a flow chart of a method for site selection and route planning of multiple distribution logistics centers of the present invention.

[0053] Figure 2 It is a flow chart of the present invention for obtaining historical express delivery volume data of each delivery terminal node in the delivery area.

[0054] Figure 3 It is a flow chart of the present invention for calculating the location of a logistics center based on the delivery distance and historical express delivery volume data of the delivery terminal nodes.

[0055] Figure 4 It is a flow chart of the present invention for calculating the basic logistics volume of a logistics center based on the fluctuation of the total distribution volume within a year.

[0056] Figure 5 It is a flowchart of the present invention which predicts the current year's express volume data based on historical express volume data using an autoregressive integral moving average model, and then calculates overflow flow data based on the current year's express volume data and basic logistics volume.

[0057] Figure 6 It is a flow chart of the present invention for obtaining data of rentable distribution logistics sites within a preset range near a logistics center location.

[0058] Figure 7 It is a flow chart of the present invention for matching the combination of leased logistics sites with the lowest cost based on overflow flow data.

[0059] Figure 8 It is a flow chart of the present invention for calculating the cost data of each candidate logistics station solution and arranging them in order from low to high cost.

[0060] Fig. 9 It is a flow chart of the present invention for dynamically allocating corresponding distribution terminal nodes based on renting logistics sites. DETAILED DESCRIPTION

[0061] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0062] See also Figures 1 to 9 The present invention provides a method for site selection and route planning of multiple distribution logistics centers, comprising:

[0063] S101 obtains historical express delivery volume data of each delivery terminal node in the delivery area;

[0064] The specific steps include:

[0065] S201 collects logistics data from different channels of each distribution terminal node;

[0066] Collect data from multiple sources, including but not limited to enterprise resource planning (ERP) systems, warehouse management systems (WMS), transportation management systems (TMS), etc. There are also data interfaces shared with partners or suppliers, such as e-commerce platforms and express service provider APIs.

[0067] During the collection process, be sure to confirm that all relevant data has been covered, especially key information that may affect the final analysis results, such as order number, timestamp, recipient address, package weight and size, etc.

[0068] S202 cleans up duplicates, outliers or incomplete records in logistics data;

[0069] Use unique identifiers (such as order numbers) to identify and remove duplicate entries to avoid statistical errors. By setting a reasonable threshold range, screen out data points that deviate significantly from normal levels and conduct further investigation. If it is confirmed to be an error, correct it; if it is caused by a special event, consider how to deal with it appropriately in subsequent analysis. For data rows that are missing key fields, try to find supplementary information from other places, or use interpolation, mean filling and other methods to make reasonable inferences. After the cleaning is completed, check the data set again to ensure that no problems are missed and that all data is accurate.

[0070] S203 formats the cleaned logistics data to obtain historical express delivery volume data.

[0071] According to the analysis requirements, the raw data is converted into a format that is easy to understand and use. For example, convert date strings into date objects, convert text descriptions into categorical variables, etc. Based on the cleaned data, express volume statistics are generated by day, week, month, or other periods. This helps to quickly view the overall trend and performance in a specific time period. In addition to basic quantity statistics, additional dimensions can be added, such as seasonal factors, promotional activities, etc., to more fully understand the reasons for changes in express volume.

[0072] S102 calculates the location of the logistics center based on the delivery distance and the historical express volume data of the delivery terminal node; the specific steps include:

[0073] S301 obtains all potential logistics center locations and obtains data of a group of centers to be selected;

[0074] Determine the geographical areas where logistics centers may be established based on business needs and service coverage. These areas can be cities, industrial areas, near transportation hubs, etc. Through market research, government public resources, commercial real estate platforms (such as LoopNet, CBRE), and cooperation with local intermediaries, collect specific information for each potential location, including:

[0075] Geographic location coordinates: latitude and longitude, used for subsequent distance calculations.

[0076] Infrastructure conditions: such as transportation convenience, power supply stability, communication facilities, etc.

[0077] Cost factors: rent, land price, construction cost, labor cost, etc.

[0078] Policy environment: tax incentives, environmental regulations, land use restrictions, etc.

[0079] Based on pre-set criteria (such as budget, scale requirements, and expansion potential), eligible candidate center group data is screened from a large number of candidate locations.

[0080] S302 calculates the transportation time from each distribution terminal node to each candidate center in the candidate center group data;

[0081] Choose the most appropriate distance or time measurement method based on the actual situation. You can choose straight-line distance (Euclidean distance), road network distance, or estimated travel time. For short-distance delivery within the city, it is recommended to use the actual road network and real-time traffic data to estimate the travel time. With the help of geographic information system (GIS) software such as ArcGIS, QGIS, etc., import the location data of all delivery terminal nodes and candidate centers, and batch calculate the distance or time between each pair of points. You can also use online API services (such as Google Maps API, Mapbox API) for accurate calculations.

[0082] S303 uses the centroid method to calculate the location of the logistics center based on transportation time and historical express delivery volume data.

[0083] Set the geographic coordinates (x_i, y_i) of the delivery terminal node, the historical express volume (w_i), and the transportation time matrix generated in the above steps.

[0084] Apply the center of gravity formula: Use the center of gravity method to calculate the optimal logistics center location. The formula is as follows:

[0085]

[0086] Where X and Y represent the optimal latitude and longitude of the logistics center, respectively; n is the number of distribution terminal nodes; w i is the historical express delivery volume of the ii-th node; x i and i are the latitude and longitude of the i-th node respectively.

[0087] In order to more accurately reflect the actual operation situation, the express delivery volume of each node can be weighted when calculating the center of gravity. The weight can be the transportation time or other relevant indicators. The modified formula is:

[0088]

[0089] Among them, t i Represents the transportation time from the ii-th node to the candidate center.

[0090] Check whether the calculated location of the logistics center is reasonable, such as whether it is within the feasible area, whether it is close to the main transportation network, etc. If necessary, fine-tune the model parameters according to the actual situation.

[0091] S103 calculates the basic logistics volume of the logistics center based on the fluctuation of the total distribution volume within a year;

[0092] The specific steps include:

[0093] S401 obtains the delivery data of all delivery terminal nodes in the previous year based on the historical express delivery volume data;

[0094] Data collection: Extract detailed delivery records of all delivery terminal nodes in the past year from the company's internal systems (such as ERP, WMS, TMS) and other related channels (such as e-commerce platform APIs, third-party logistics service providers).

[0095] Identify a specific time period as the reference year, usually a full calendar year (e.g., January 1 to December 31 of the previous year) to ensure that all seasonal and cyclical changes are covered. Ensure that the collected data is complete, especially key fields such as order date, quantity, weight, destination, etc. For missing or incomplete records, take appropriate filling or deletion measures.

[0096] S402 obtains holiday sales data based on the delivery data;

[0097] List all holidays that affect your business, including statutory holidays, traditional holidays, and company-specific promotional days. These days tend to result in a significant increase in delivery volume. Separate the specific delivery volume data for each holiday from the annual delivery data. Be careful to distinguish between different types of holidays, as they may have different effects on delivery volume.

[0098] S403 obtains the median of the holiday express delivery volume data to replace all holiday express delivery volume data;

[0099] For each identified holiday type, calculate the median of its delivery volume data. Using the median instead of the mean can reduce the impact of extreme values ​​and obtain more robust results. Replace each value in the original holiday delivery volume data with the calculated median. This is done to eliminate short-term fluctuations caused by holidays so that the data can better reflect normal business conditions. Merge the replaced data with non-holiday data to form a new, smoother full-year delivery volume data set. This step helps to better identify long-term trends and cyclical patterns in subsequent analysis.

[0100] S404 calculates the basic logistics volume for each selected time period throughout the year based on the replaced holiday express delivery volume data.

[0101] Divide the year into several reasonable time periods, such as months, quarters, or other custom periods, based on business characteristics and management needs. Make sure each time period is large enough to capture meaningful trends, but not too broad to obscure details.

[0102] For each time period, calculate the average volume. If you want a more robust result, you can choose the median volume for that time period instead of the average. If there is a clear long-term growth or decline trend, you can apply an appropriate trend adjustment method before calculating the average, such as a moving average or exponential smoothing method. Draw a basic logistics volume curve over time to intuitively show the expected demand level for each time period. This curve will become an important basis for future planning and decision-making.

[0103] Through the above detailed steps, the basic logistics volume of the logistics center can be scientifically and reasonably calculated based on the fluctuation of the total distribution volume within a year. This method can not only help predict future logistics demand, but also guide important decisions such as resource allocation and facility expansion, thereby improving operational efficiency and service quality.

[0104] S104 predicts the current year's express volume data based on historical express volume data using an autoregressive integrated moving average model, and then calculates overflow logistics volume data based on the current year's express volume data and basic logistics volume;

[0105] The specific steps include:

[0106] S501 uses the ADF test to determine whether the time series needs to be differencing to make it stationary;

[0107] Choose an appropriate test method: Use the Augmented Dickey-Fuller (ADF) test or other unit root tests (such as the KPSS test) to check whether the time series is a stationary process.

[0108] Perform ADF test: usually set at 0.05 or more stringent level. If the p-value is less than the set significance level, the null hypothesis (i.e., the existence of a unit root) is rejected and the time series is considered to be stationary; otherwise, the null hypothesis is accepted, indicating that differential processing is required. Based on the results of the ADF test, determine how many differential operations are required to make the time series reach a stable state. Generally, the number of differential operations will not exceed 2.

[0109] S502 uses the autocorrelation function graph and the partial autocorrelation function graph to estimate p and q values ​​to construct an ARIMA model;

[0110] Use the Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) plots to visually observe the correlations at different lag orders. The tailing phenomenon in the ACF plot and the location of the truncation in the PACF plot can help determine the p-value. The tailing phenomenon in the PACF plot and the location of the truncation in the ACF plot can help determine the q-value. If there is a clear seasonal pattern, it is also necessary to consider the seasonal ARIMA (SARIMA) model and adjust the parameter estimation method accordingly.

[0111] S503 uses historical express delivery volume data to train the ARIMA model to ensure that the model can capture the trend, seasonality, and cyclical components in the data;

[0112] Divide the historical express delivery volume data into a training set and a validation set. The training set is used for model fitting, and the validation set is used to evaluate model performance. Initialize the ARIMA model using the selected (p, d, q) parameters and fit it using methods such as maximum likelihood estimation. Check whether the residuals are white noise (i.e., there is no obvious pattern) to ensure that the model has fully captured the structured information in the data. Based on the model diagnosis results, appropriately adjust the parameters or introduce exogenous variables (such as holiday effects, promotional activities, etc.) to further improve the model's fit and predictive ability.

[0113] S504 evaluates the prediction accuracy of the model through cross validation;

[0114] Use methods such as Leave-One-Out and k-fold cross validation to train and test the model multiple times to obtain multiple error estimates. Common evaluation indicators include mean square error (MSE), root mean square error (RMSE), mean absolute error (MAE), etc. You can also consider using R 2 The score measures the proportion of data variability explained by the model. If you try multiple configurations or different modeling methods, you can compare their prediction accuracy and choose the best performing model as the final version.

[0115] S505 uses the trained ARIMA model to predict the express delivery volume of the current year on a monthly basis to obtain the predicted logistics volume;

[0116] Use the trained ARIMA model to predict the express delivery volume for each month in the next year and output the prediction results. Create a line chart or bar chart to show the predicted express delivery volume trend for intuitive understanding.

[0117] S506 When the predicted logistics volume exceeds the basic logistics volume, the excess is the overflow logistics volume.

[0118] Recall the previously calculated base volume, which is a steady demand level based on historical data during off-peak periods. For each predicted time point, if the predicted volume is greater than the base volume, the difference between the two is the overflow volume for that time period. With time as the horizontal axis and overflow volume as the vertical axis, a curve of overflow volume over time is plotted. This helps identify peak periods and points in time when additional resources are needed.

[0119] S105: obtaining data of leasable distribution logistics sites within a preset range near the logistics center location, wherein the logistics site data includes area price and facility type;

[0120] The specific steps include:

[0121] S601 sets the search radius based on the location of the logistics center;

[0122] Set a reasonable search radius based on business needs and service coverage. This radius should be large enough to cover enough candidate locations, but not too broad to be difficult to manage. The selected search radius should take into account the layout of the transportation network, such as proximity to major roads, railway stations or other transportation hubs to optimize transportation routes and times. Combined with local real estate market conditions, such as supply and demand, rental levels, etc., to adjust the search radius to ensure that you find a cost-effective logistics site.

[0123] S602 maps the logistics center location and search radius to a logistics real estate database for search;

[0124] Choose the right database: Use professional logistics real estate database services, such as CBRE, JLL (Jones LangLaSalle), LoopNet, or commercial real estate platforms. These platforms usually provide detailed logistics site information, including location, area, price, facility type, etc.

[0125] Enter the specific latitude and longitude of the logistics center into the database search interface. Specify the previously determined search radius in the search tool. Add additional screening conditions as needed, such as minimum / maximum available area, expected rental range, specific facility requirements (freezers, dry warehouses, loading and unloading platforms, etc.). Start the search function and let the system automatically retrieve all logistics sites that meet the conditions and generate a preliminary list of results.

[0126] S603 lists all candidate sites that meet the conditions based on the searched information, and obtains data on the distribution logistics sites that can be rented.

[0127] Extract key information from the data returned by the database and create a structured table or document to record the details of each candidate site, including but not limited to:

[0128] Location: Precise address and description of surroundings.

[0129] Square footage: The total area and specific size of the space available for lease.

[0130] Price: Monthly or annual rent per square meter, plus other possible costs (property management fees, utilities, etc.).

[0131] Facility type: infrastructure and services provided, such as warehouse type (ordinary warehouse, cold storage), loading and unloading facilities, security measures, etc.

[0132] Transportation accessibility: Assess the ease of reaching key customer groups or suppliers.

[0133] Eventually, a list of all eligible candidate sites will be formed, which will serve as an important reference for subsequent decision-making.

[0134] S106 matches the combination of rental logistics sites with the lowest cost based on overflow flow data;

[0135] The specific steps include:

[0136] S701 obtains overflow flow data;

[0137] Obtain the current year's express delivery volume forecast data from the previous ARIMA model or other forecasting method. Refer to the previously calculated base volume, which is a stable demand level based on historical data during off-peak periods. For each predicted time point, if the predicted volume is greater than the base volume, the difference between the two is the overflow volume during that time period. With time as the horizontal axis and overflow volume as the vertical axis, draw a curve of overflow volume over time to intuitively identify peak periods and time points that require additional resource support.

[0138] S702 calculates the amount and type of additional storage required in each time period based on the overflow flow data;

[0139] Divide the year into reasonable time periods (e.g., months, quarters) and analyze the overflow flow for each time period separately. Estimate the amount of additional storage space required for each time period based on the overflow flow. Also identify the specific facility type, given that different types of goods may have different storage requirements (e.g., freezer, warehouse). Identify if there are obvious seasonal or cyclical patterns and adjust the estimate accordingly to ensure that it more accurately reflects actual needs.

[0140] S703: matching the data of the available distribution logistics sites based on the additional storage volume and type to obtain a logistics site solution to be selected;

[0141] From the previously collected data on available distribution logistics sites, select candidate sites that can meet additional warehousing needs (including area and facility type). Prioritize locations close to major customer groups or suppliers to reduce transportation time and costs. Consider the possibility of future business growth and select sites with expansion potential. Generate a list of all eligible candidate sites for each time period as an important reference for subsequent decision-making.

[0142] S704 calculates the cost data of each candidate logistics station solution and arranges them in order from low to high according to cost.

[0143] The specific steps include:

[0144] S801 collects various cost data for each candidate logistics site;

[0145] Cost data includes the following items:

[0146] Rent: Fixed rent per month or year.

[0147] Operating expenses: such as water and electricity charges, property management fees, etc.

[0148] Maintenance costs: daily repair and upkeep costs.

[0149] Labor costs: employee wages, benefits, etc.

[0150] Transportation cost: transportation cost from logistics station to each distribution point.

[0151] Equipment costs: Purchase or lease of required handling equipment, storage facilities, etc.

[0152] Insurance costs: property insurance, liability insurance, etc.

[0153] Tax and legal costs: local taxes, registration fees, license fees, etc.

[0154] Other potential costs: such as environmental protection measures, safety measures, etc.

[0155] S802 converts different types of costs in the cost data into a unified unit;

[0156] Convert all costs to the same time unit (such as monthly or annual) to facilitate comparison. Make sure all cost data is in the same currency unit, and make exchange rate adjustments if necessary.

[0157] S803 calculates fixed costs and variable costs based on the converted data to obtain the logistics costs to be selected;

[0158] Fixed costs are those costs that do not change with changes in business volume, such as rent, equipment rental fees under long-term contracts, etc. Variable costs are costs that increase with the increase in business volume, such as transportation costs, labor costs priced by piece, etc. Add the fixed costs and variable costs together to get the total cost of each candidate site.

[0159] S804 arranges the logistics costs to be selected from low to high.

[0160] Based on the calculated total cost, rank all candidate sites, starting with the lowest cost. Test the impact of key variables (such as rent increases, shipping rate changes) on the total cost to ensure that your cost estimates are flexible.

[0161] S107 dynamically allocates corresponding distribution terminal nodes based on the leased logistics sites.

[0162] The specific steps include:

[0163] S901 calculates the actual road distance from each distribution terminal node to each logistics site;

[0164] Ensure that you have the precise geographic coordinates (latitude and longitude) of all distribution terminal nodes and leased logistics sites. This data can be obtained from internal enterprise systems, third-party mapping services, or field measurements. Using the real road network to calculate the shortest path distance between two points is closer to the actual situation.

[0165] S902 selects the nearest logistics station for delivery.

[0166] Based on the calculation results in S901, a transportation matrix is ​​created, in which each row represents a distribution terminal node, each column represents a logistics site, and the matrix elements represent the distance or travel time from the node to the corresponding site. By default, the nearest logistics site should be given priority for distribution to reduce transportation time and cost. In order to avoid overloading some sites while others are idle, a load balancing mechanism can be introduced during allocation, taking into account the current workload and service capacity of each site.

[0167] Through the above detailed steps, you can dynamically allocate corresponding delivery terminal nodes based on the leased logistics sites, thereby improving delivery efficiency, reducing costs and improving customer satisfaction. This approach not only helps to optimize current operations, but also provides a solid foundation for future expansion and development.

[0168] What is disclosed above is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. A method for site selection and route planning of multiple distribution logistics centers. It is characterized in that Including: obtaining historical express delivery volume data of each delivery terminal node in the delivery area; Calculate the location of the logistics center based on the delivery distance and historical express volume data of the delivery terminal nodes; Calculate the basic logistics volume of the logistics center based on the fluctuation of the total distribution volume within a year; Based on the historical express volume data, the autoregressive integrated moving average model is used to predict the current year's express volume data, and then the overflow logistics volume data is calculated based on the current year's express volume data and the basic logistics volume; Obtaining data on leasable distribution logistics sites within a preset range near the logistics center location, the logistics site data including area price and facility type; Match the combination of rental logistics sites with the lowest cost based on overflow flow data; Dynamically allocate corresponding distribution terminal nodes based on rented logistics sites.

2. A method for site selection and route planning of multiple distribution logistics centers as claimed in claim 1, characterized in that: The specific steps of obtaining the historical express delivery volume data of each delivery terminal node in the delivery area include: Collect logistics data from different channels of each distribution terminal node; Clean up duplicates, outliers, or incomplete records in logistics data; Format the cleaned logistics data to obtain historical express delivery volume data.

3. A method for site selection and route planning of multiple distribution logistics centers as claimed in claim 2, characterized in that: The specific steps of calculating the location of the logistics center based on the delivery distance and the historical express volume data of the delivery terminal node include: Obtain all potential logistics center locations and obtain the data of the candidate center group; Calculate the transportation time from each distribution terminal node to each candidate center in the candidate center group data; The location of logistics centers is calculated using the centroid method based on transportation time and historical express volume data.

4. A method for site selection and route planning of multiple distribution logistics centers as claimed in claim 3, characterized in that: The specific steps of calculating the basic logistics volume of the logistics center based on the fluctuation of the total distribution volume within a year include: Obtain the delivery data of all delivery terminal nodes in the previous year based on historical express delivery volume data; Get holiday sales data based on delivery data; Get the median of holiday express delivery volume data to replace all holiday express delivery volume data; The basic logistics volume is calculated for each selected time period throughout the year based on the replaced holiday express volume data.

5. A method for site selection and route planning of multiple distribution logistics centers as claimed in claim 4, characterized in that: The specific steps of predicting the current year's express volume data based on the historical express volume data using the autoregressive integrated moving average model, and then calculating the overflow flow data based on the current year's express volume data and the basic logistics volume include: Use the ADF test to determine whether the time series needs to be differentiated to make it stationary; Use autocorrelation and partial autocorrelation plots to estimate p and q values ​​to build an ARIMA model; Train the ARIMA model using historical express delivery volume data to ensure that the model can capture the trend, seasonality, and cyclical components in the data; Evaluate the prediction accuracy of the model through cross-validation; Use the trained ARIMA model to predict the express delivery volume of the current year on a monthly basis to obtain the predicted logistics volume; When the predicted logistics volume exceeds the basic logistics volume, the excess is the overflow logistics volume.

6. A method for site selection and route planning of multiple distribution logistics centers as claimed in claim 5, characterized in that: The specific steps of obtaining the data of the leasable distribution logistics sites within a preset range near the logistics center location include: Set the search radius based on the location of the logistics center; Map the logistics center location and search radius to the logistics real estate database for search; Based on the information obtained from the search, all eligible candidate sites are listed to obtain the data of rentable distribution logistics sites.

7. A method for site selection and route planning of multiple distribution logistics centers as claimed in claim 6, characterized in that: The specific steps of matching the combination of leased logistics sites with the lowest cost based on the overflow flow data include: Obtain spill flow data; Calculate the amount and type of additional storage required at various time periods based on spill flow data; Based on the additional storage volume and type, matching is performed in the data of available distribution logistics sites to obtain the logistics site plan for selection; Calculate the cost data of each candidate logistics station plan and arrange them in order from low to high cost.

8. A method for site selection and route planning of multiple distribution logistics centers as claimed in claim 7, characterized in that: The specific steps of calculating the cost data of each candidate logistics station solution and arranging them in order from low to high cost include: For each candidate logistics site, collect various cost data; Convert different types of costs in cost data into uniform units; Calculate fixed costs and variable costs based on the converted data to obtain the logistics costs to be selected; Arrange the logistics costs to be selected from low to high.

9. A method for site selection and route planning of multiple distribution logistics centers as claimed in claim 8, characterized in that: The specific steps of dynamically allocating corresponding distribution terminal nodes based on the leased logistics site include: Calculate the actual road distance from each distribution terminal node to each logistics site; Choose the nearest logistics station for delivery.

Images