Inventory control method, system and equipment based on intelligent algorithm and medium

Abstract

The invention discloses an inventory control method, system and device based on an intelligent algorithm and a medium. The method comprises the steps of collecting storage multi-dimensional data; inputting the multi-dimensional data into a prediction model, obtaining warehouse stock characteristics through the prediction model, and obtaining a predicted value of the material consumption according to the stock characteristics; according to the predicted value of the material consumption, combined with constraint conditions, calculating a dynamic safety coefficient of storage; and designing a target function based on the dynamic safety coefficient, solving the target function through a genetic algorithm, and obtaining the optimal safe inventory level of storage. According to the invention, through double analysis and weighted prediction of the time sequence data and the external influence factors, the accuracy of material consumption prediction can be improved; the adaptability of an inventory strategy can be enhanced by combining a dynamic safety coefficient; and multi-dimensional constraint conditions are set and parameters are optimized, so that multi-aspect interest demands of inventory holding cost, storage capacity, service level, replenishment frequency and the like are balanced, and the inventory cost is effectively reduced.

Description

Technical Field

[0001] This invention relates to the field of warehouse management technology, and in particular to an inventory control method, system, equipment and medium based on intelligent algorithms. Background Technology

[0002] In modern supply chain management, safety stock is a crucial buffer mechanism for enterprises to cope with demand fluctuations and supply uncertainties, and its proper setting is of great significance. On the one hand, it can effectively cope with demand fluctuations and forecasting errors, reduce the risk of supply chain disruptions, and ensure customer service levels; on the other hand, it helps balance inventory costs and stockout costs, directly impacting the company's operational efficiency and customer satisfaction. In recent years, the process of global economic integration has accelerated, and the market environment has become increasingly complex. The diversification and rapid changes in consumer demand have made market demand fluctuations more difficult to predict; at the same time, suppliers are affected by various factors such as natural disasters, political situations, and raw material shortages, significantly increasing supply uncertainty. This has brought unprecedented challenges to enterprises' safety stock management.

[0003] While the development of emerging technologies such as big data and artificial intelligence has provided new approaches to solving safety stock management problems, the market currently lacks solutions that can effectively integrate multi-source data and achieve intelligent dynamic adjustment of safety stock. Most existing inventory management systems employ static inventory strategies, which lack flexibility and struggle to adapt to rapidly changing market environments. For example, in the face of sudden increases in market demand or delayed deliveries from suppliers, static inventory strategies cannot adjust inventory levels in a timely manner, easily leading to stockouts or inventory buildup, thereby increasing operating costs and reducing customer satisfaction. Therefore, developing an algorithm that can dynamically adapt to market changes and automatically optimize inventory levels has become a crucial problem urgently needing to be solved in the field of supply chain management. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an inventory control method based on intelligent algorithms to solve the problem that traditional inventory management methods cannot effectively integrate multi-source data and achieve intelligent dynamic adjustment of safety stock in complex and ever-changing market environments, thus making it difficult to balance inventory costs and service levels and adapt to rapid market changes.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides an inventory control method based on intelligent algorithms, comprising:

[0008] Collect multi-dimensional data on warehousing;

[0009] The multi-dimensional data is input into the prediction model, the prediction model is used to obtain the warehouse inventory characteristics, and the predicted value of material consumption is obtained based on the inventory characteristics.

[0010] Based on the predicted consumption of the materials and the constraints, the dynamic safety factor of the warehouse is calculated.

[0011] Based on the dynamic safety factor, an objective function is designed, and the objective function is solved using a genetic algorithm to obtain the optimal safety stock level for warehousing.

[0012] As a preferred embodiment of the intelligent algorithm-based inventory control method of the present invention, the method includes: inputting the multi-dimensional data into a prediction model, obtaining warehouse inventory characteristics through the prediction model, and obtaining predicted values ​​of material consumption based on the inventory characteristics, including:

[0013] Analyze the time series data in the multi-dimensional data to obtain the trend and seasonal characteristics of material demand, and obtain a first predicted value based on the trend and seasonal characteristics;

[0014] Analyze the external influencing factors in the multi-dimensional data, establish a regression equation, and obtain the second predicted value;

[0015] The first and second predicted values ​​are weighted and calculated to obtain the predicted value of material consumption.

[0016] The beneficial effects of this preferred technical solution are as follows: by analyzing time series data and external influencing factor data respectively, key factors affecting material consumption are explored from different perspectives; the two predicted values ​​are weighted and calculated, which combines the advantages of the two prediction methods and is more adaptable to the complex and ever-changing market environment than a single prediction method, thus improving the accuracy of material consumption prediction.

[0017] As a preferred embodiment of the intelligent algorithm-based inventory control method described in this invention, the dynamic safety factor of the warehouse is calculated based on the predicted value of material consumption and in conjunction with constraints, including:

[0018] The constraints include the target service level and the volatility of material demand;

[0019] Based on the predicted value of the material consumption, obtain the prediction mean absolute error;

[0020] Based on the target service level, the volatility of material demand, and the average absolute error of the forecast, the dynamic safety factor of the warehouse is calculated.

[0021] The beneficial effects of this preferred technical solution are as follows: by combining multiple factors to calculate the dynamic safety factor, the calculation of the safety factor becomes more scientific and reasonable, and it can be dynamically adjusted according to the actual business situation. This provides key and accurate parameters for determining the optimal safety stock level, thereby enhancing the adaptability and reliability of the inventory strategy.

[0022] As a preferred embodiment of the intelligent algorithm-based inventory control method of the present invention, the method includes: designing an objective function based on the dynamic safety factor, solving the objective function using a genetic algorithm, and obtaining the optimal safety stock level for warehousing, including:

[0023] Based on the replenishment lead time and demand standard deviation of acquired materials, and combined with the aforementioned dynamic safety factor, an objective function is designed.

[0024] Set up a genetic algorithm to solve the constraints, and then use the genetic algorithm to solve the objective function.

[0025] Based on total cost, service level, and replenishment frequency, a comprehensive performance index value is calculated, and the genetic algorithm parameters are optimized and adjusted based on the comprehensive performance index value.

[0026] The beneficial effects of this preferred technical solution are as follows: it comprehensively considers factors such as supply chain response speed, demand uncertainty, and market and forecasting factors, allowing the objective function to more accurately reflect the actual inventory management needs. It sets constraints to avoid problems such as excessively high inventory costs, insufficient storage space, declining service levels, or overly frequent replenishment. This results in more efficient acquisition of the optimal safety stock level in warehousing, achieving a multi-objective balance and overall efficiency improvement for the enterprise in inventory management.

[0027] As a preferred embodiment of the intelligent algorithm-based inventory control method described in this invention, the genetic algorithm solves for constraints including inventory holding cost constraints, storage capacity constraints, service level constraints, and replenishment frequency constraints.

[0028] As a preferred embodiment of the intelligent algorithm-based inventory control method described in this invention, the objective function is expressed as:

[0029]

[0030] Where S(t) is the optimal safety stock level, σ(t) is the standard deviation of demand, k(t) is the dynamic safety factor, and L is the replenishment lead time.

[0031] As a preferred embodiment of the intelligent algorithm-based inventory control method described in this invention, the formula for calculating the dynamic safety factor of the warehouse is:

[0032] k(t)=α·CSL+β·MAE(t)+γ·VOL(t)

[0033] Where k(t) is the dynamic safety factor, CSL is the target service level, MAE(t) is the mean absolute error, VOL(t) is the demand volatility, and α, β, and γ are weighting coefficients.

[0034] Secondly, the present invention provides an inventory control system based on intelligent algorithms, including: a data acquisition module for collecting multi-dimensional warehouse data;

[0035] The prediction module is used to input the multi-dimensional data into the prediction model, obtain the warehouse inventory characteristics through the prediction model, and obtain the predicted value of material consumption based on the inventory characteristics.

[0036] The calculation module is used to calculate the dynamic safety factor of the warehouse based on the predicted value of the material consumption and the constraints; based on the dynamic safety factor, an objective function is designed, and the objective function is solved by a genetic algorithm to obtain the optimal safety stock level of the warehouse.

[0037] Thirdly, the present invention provides an electronic device, comprising:

[0038] Memory and processor;

[0039] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of an inventory control method based on intelligent algorithms.

[0040] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the intelligent algorithm-based inventory control method.

[0041] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention collects multi-dimensional data and improves the accuracy of material consumption forecasting through dual analysis and weighted prediction of time series data and external influencing factors; it calculates a dynamic safety factor, enabling the safety stock level to be dynamically adjusted according to actual business conditions, thereby enhancing the adaptability of inventory strategies; and it sets multi-dimensional constraints and optimizes parameters in the genetic algorithm, balancing the interests of inventory holding costs, storage capacity, service levels, and replenishment frequency, effectively reducing inventory costs, improving service levels, and achieving multi-objective optimization in enterprise inventory management. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of the overall process of an inventory control method based on intelligent algorithms according to an embodiment of the present invention. Detailed Implementation

[0044] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0045] Example 1, referring to Figure 1 As one embodiment of the present invention, an inventory control method based on intelligent algorithms is provided, comprising:

[0046] S100: Collects multi-dimensional data on warehousing;

[0047] S102: Input multi-dimensional data into the prediction model, obtain warehouse inventory characteristics through the prediction model, and obtain the predicted value of material consumption based on the inventory characteristics;

[0048] S104: Calculate the dynamic safety factor of the warehouse based on the predicted value of material consumption and the constraints.

[0049] S106: Based on the dynamic safety factor, design the objective function, solve the objective function through a genetic algorithm, and obtain the optimal safety stock level for warehousing.

[0050] It should be noted that in a complex market environment, a single inventory management approach is insufficient to adapt to changes in demand and supply. This invention collects multi-dimensional warehouse data, enabling comprehensive acquisition of inventory-related information and providing a solid basis for subsequent analysis. By inputting this multi-dimensional data into a prediction model, combining the trend and seasonal characteristics of time-series data with external influencing factors, and employing dual analysis and weighted calculations, the predicted value of material consumption is obtained, improving prediction accuracy and providing reliable support for inventory decisions. Based on the predicted value, a dynamic safety factor is calculated using the target service level, material demand volatility, and the mean absolute error of the prediction, allowing for flexible adjustments to safety stock according to actual business conditions. An objective function is designed based on the dynamic safety factor and solved using a genetic algorithm under constraints such as inventory holding costs, warehousing capacity, service level, and replenishment frequency, thereby obtaining the optimal safety stock level and achieving a balance between inventory costs and service levels while satisfying various business conditions. This enhances the intelligence and dynamic adaptability of inventory management and reduces inventory costs.

[0051] Example 2, refer to Figure 1 This is one embodiment of the present invention. Based on the above embodiment, an inventory control method based on intelligent algorithms is provided.

[0052] In this embodiment of the invention, step S100 involves collecting multi-dimensional warehousing data, including historical demand data, inventory status data, supplier delivery data, and market fluctuation indicators.

[0053] For example, historical demand data D t This represents the quantity demanded at time t, where t is measured in days. For example, D1 represents the quantity demanded on day 1, D2 represents the quantity demanded on day 2, and so on.

[0054] Inventory status data includes: Current inventory level: I c Indicates the current inventory quantity; Inventory limit: I max Indicates the maximum allowable inventory level; lower inventory limit: I min This indicates the minimum quantity of inventory that cannot be lowered, and is generally associated with safety stock.

[0055] Supplier delivery data L i This represents the time interval between placing an order with a supplier and receiving the goods, where i represents different suppliers. Delivery quantity Q i,t This represents the quantity of goods received from supplier i at time t.

[0056] Market demand volatility coefficient σ D Used to measure the degree of fluctuation in market demand, it can be calculated using methods such as the standard deviation of historical demand data. A larger σ... DA large fluctuation in market demand indicates that market demand is relatively stable, while a small fluctuation indicates that market demand is relatively stable.

[0057] In this embodiment of the invention, the preprocessing of multi-dimensional data includes the following steps:

[0058] Data cleaning methods, such as moving averages and exponential smoothing, are used to remove abnormal fluctuations and noise. Moving averages smooth data by calculating the average value over a specific time window, thus removing short-term abnormal fluctuations and noise. For time series data D... i The formula for calculating the k-period moving average is:

[0059]

[0060] Among them, MA t Let D represent the moving average over time t, where k is the window size of the moving average, and D is the moving average value. i This represents the original demand data for time i. The formula calculates the average of data from the past k time points to obtain a smoothed value for the current time point, thereby reducing random fluctuations in the data.

[0061] Exponential smoothing assigns higher weight to recent data, better reflecting the latest trends in data changes. The calculation formula is as follows:

[0062] ES t =αD i +(1-α)ES t-1

[0063] Among them, ES t D represents the exponential smoothing value of time t, where α is the smoothing coefficient, and its value ranges from 0 < α < 1. i It is the raw demand data for time i, ES t-1 It is the exponentially smoothed value at time t-1. This formula shows that the current smoothed value is a weighted average of the current original data and the smoothed value of the previous period, with the weights determined by the smoothing coefficient α.

[0064] Multiple imputation algorithms are used to handle missing values ​​to ensure data continuity; linear regression imputation is employed, with the number of missing values ​​R at imputation time t being [missing value]. t A linear regression model can be established:

[0065] R t =β0+β1x 1,t +β2x 2,t +...+β m x m,t +∈ t

[0066] Where, x i,t β is the value of the i-th characteristic variable related to demand at time t.i It is the regression coefficient, ∈ t This is the error term.

[0067] Statistical anomaly detection algorithms: The Z-score method identifies outliers by calculating the degree of deviation of data points from the mean. For data point D... t The formula for calculating its Z-score is:

[0068]

[0069] Where μ is the mean of the data, and σ is the standard deviation of the data. Typically, when Z... t When >3, it is considered that D t It is an outlier.

[0070] All data is standardized, and a common standardization method is Z-score standardization. For each element x in the historical demand data matrix x(t), ... ij The standardized formula is:

[0071]

[0072] Where, μ j σ is the mean of the j-th data point. j It is the standard deviation of the j-th data point, and the standardized data. It has the characteristics of a mean of 0 and a standard deviation of 1.

[0073] The final standardized historical demand data matrix x * The elements in (t) are The matrix has an dimension of n×m, where n is the number of time series samples and m is the number of data dimensions.

[0074] In one alternative implementation, the multi-dimensional data may also include macroeconomic data, competitor inventory data, promotional activity arrangement data, etc.

[0075] In one optional implementation, preprocessing can include: using wavelet transform for data denoising, leveraging the multi-resolution analysis characteristics of wavelet functions to decompose the data into different frequency bands, removing high-frequency noise while retaining the main features of the data. For missing value handling, in addition to linear regression imputation, the K-nearest neighbor algorithm can also be used. For outlier detection, density-based spatial clustering algorithms can be used, which can identify data points in low-density regions as outliers based on the density distribution of data points.

[0076] In this embodiment of the invention, step S102 involves inputting multi-dimensional data into a prediction model, obtaining warehouse inventory characteristics through the prediction model, and obtaining predicted values ​​of material consumption based on the inventory characteristics. The method also includes sub-steps A1-A3:

[0077] A1: Analyze the time series data in the multi-dimensional data to obtain the trend and seasonal characteristics of material demand, and obtain the first forecast value based on the trend and seasonal characteristics;

[0078] A2: Analyze the external influencing factors in the multi-dimensional data, establish a regression equation, and obtain the second predicted value;

[0079] A3: Weight the first and second predicted values ​​to obtain the predicted value of material consumption.

[0080] In this embodiment of the invention, obtaining the trend and seasonal characteristics of material demand specifically includes: using a seasonal ARIMA model to obtain the trend and seasonal characteristics of material demand. The formula for the seasonal ARIMA model is:

[0081]

[0082] Where B is the lag operator, i.e., By t =y t-1 B s y t =y t-s , It is a non-seasonal autoregressive polynomial, ω(B s ) is a seasonal autoregressive polynomial, θ(B) is a non-seasonal moving average polynomial, and θ(B) is a seasonal moving average polynomial. t It is a white noise sequence.

[0083] In this embodiment of the invention, the consumption of materials is not only affected by historical data, but may also be affected by external factors such as temperature T. t GDP growth rate t Driven by factors such as [various factors], a multiple linear regression model is added for supplementary prediction.

[0084] y t =β0+β1T t +β2G t +...+β k X kt +∈ t

[0085] Among them, y t β is the amount of resources consumed at time t, β0 is the intercept term, and β i (i = 1, 2, 3, ..., k) are regression coefficients, representing the weight of each factor on the amount of material consumption, Xkt It is the value of the k-th influencing factor at time t; ∈ t This is the error term.

[0086] Estimating the regression coefficient β using the least squares method i After obtaining the regression equation, the predicted values ​​X of future influencing factors are then used. k,t+h Calculate the predicted values ​​of the multiple linear regression model. Represented as:

[0087]

[0088] In this embodiment of the invention, the predicted value of the final material consumption... Predicted values ​​from the seasonal ARIMA model and the predicted values ​​of the multiple linear regression model It is composed of weighted combinations and is represented as:

[0089]

[0090] Where λ is the weighting coefficient.

[0091] In one alternative implementation, the first predicted value can be obtained using the Prophet model. Specifically, the time series data from the multi-dimensional data is organized into the format required by the Prophet model, containing two columns: date and quantity demanded. Then, the data is fitted. The Prophet model automatically performs trend decomposition, breaking down the time series into trend components, seasonal components, and holiday components. During the fitting process, the model automatically adjusts its parameters based on the characteristics of the data to better capture the trend and seasonal features. After fitting is complete, by setting the prediction time range, the Prophet model can generate the first predicted value.

[0092] In one alternative implementation, the second predicted value can be obtained using a support vector regression model. Specifically, for external influencing factor data in multi-dimensional data, such as temperature Tt and GDP growth rate Gt, the data is divided into training and testing sets. During the training phase, a suitable kernel function is selected to construct the support vector regression model. The goal of the support vector regression model is to find an optimal hyperplane such that the distance from the training data points to the hyperplane is minimized, while ensuring that the prediction error is within a certain tolerance range. By adjusting the parameters of the support vector regression model, the model is trained using the training set to obtain a model that can well fit the relationship between external influencing factors and material consumption. During the testing phase, the predicted values ​​of future external influencing factors are input into the trained support vector regression model, and the model output is the second predicted value.

[0093] It should be noted that this invention utilizes a seasonal ARIMA model for in-depth analysis of time-series data, accurately capturing the trends and seasonal characteristics of material demand over time, effectively solving the problem of traditional methods struggling to cope with cyclical changes. Meanwhile, the multiple linear regression model incorporates external influencing factors such as temperature and GDP growth rate into the prediction system, overcoming the limitations of relying solely on historical data. This quantifies the driving effect of external variables on material consumption, enhancing the comprehensiveness of the prediction. By weighting and combining the predictions from both models, the advantages of both are fully utilized. It considers both the time-series characteristics of the data and the influence of the external environment. Compared to single-model prediction, this improves the accuracy and stability of material consumption prediction, making the prediction results more closely reflect the complex and ever-changing market reality. This provides a reliable foundation for subsequent dynamic safety factor calculation and safety stock optimization, reducing inventory management risks.

[0094] In this embodiment of the invention, step S104, which calculates the dynamic safety factor of the warehouse based on the predicted value of material consumption and in conjunction with the constraints, further includes sub-steps B1-B3:

[0095] B1: Constraints include target service level and material demand volatility;

[0096] B2: Obtain the average absolute error of the forecast based on the predicted value of material consumption;

[0097] B3: Calculate the dynamic safety factor of the warehouse based on the target service level, the volatility of material demand, and the average absolute error of the forecast.

[0098] In this embodiment of the invention, the formula for calculating the dynamic safety factor of the warehouse is:

[0099] k(t)=α·CSL+β·MAE(t)+γ·VOL(t)

[0100] Where k(t) is the dynamic safety factor, CSL is the target service level, MAE(t) is the mean absolute error, VOL(t) is the demand volatility, and α, β, and γ are weighting coefficients.

[0101] In one alternative implementation, parameters such as market supply-demand ratio, supplier on-time delivery rate, and inventory turnover rate can be used to calculate the dynamic safety factor of warehousing.

[0102] In another alternative implementation, parameters such as the economic prosperity index, the proportion of customer order urgency, and the transportation risk index can be used to calculate the dynamic safety factor of the warehouse.

[0103] In this embodiment of the invention, step S106, which designs an objective function based on a dynamic safety factor and solves the objective function using a genetic algorithm to obtain the optimal safety stock level for warehousing, also includes sub-steps C1-C3:

[0104] C1: Design an objective function based on the replenishment lead time and demand standard deviation of acquired materials, combined with a dynamic safety factor;

[0105] C2: Set the genetic algorithm to solve the constraints and solve the objective function using the genetic algorithm;

[0106] C3: Calculate the comprehensive performance index value based on total cost, service level and replenishment frequency indicators, and optimize and adjust the genetic algorithm parameters based on the comprehensive performance index value.

[0107] In this embodiment of the invention, the constraints solved by the genetic algorithm include inventory holding cost constraints, storage capacity constraints, service level constraints, and replenishment frequency constraints.

[0108] In this embodiment of the invention, the objective function is expressed as:

[0109]

[0110] Where S(t) is the optimal safety stock level, σ(t) is the standard deviation of demand, k(t) is the dynamic safety factor, and L is the replenishment lead time.

[0111] In this embodiment of the invention, the genetic algorithm solves for the constraints including:

[0112] ① Inventory holding cost constraint:

[0113] Let the total inventory holding cost be C. holding The inventory holding cost constraint can then be expressed as:

[0114] C holding =h×S(t)≤C budget

[0115] The total inventory holding cost, calculated by multiplying the unit inventory holding cost h by the safety stock level S(t), cannot exceed the pre-set cost budget ceiling C. budget .

[0116] ② Storage capacity constraints:

[0117] Taking into account the storage capacity limitations of actual warehousing facilities, the safety stock level cannot exceed the storage capacity. The formula for the storage capacity constraint is:

[0118] S(t)≤V

[0119] This formula ensures that at any time t, the optimal safety stock level S(t) will not exceed the maximum stock quantity V that the storage facility can accommodate.

[0120] ③ Service level constraints:

[0121] The service level constraint can be expressed as follows, based on the relationship between safety stock, demand standard deviation, and replenishment lead time, assuming demand follows a normal distribution:

[0122]

[0123] in, It is the cumulative distribution function of the standard normal distribution, and the safety stock level S(t) is the standard deviation of demand over the replenishment lead time. The proportion, calculated using the cumulative distribution function of the standard normal distribution, must have a probability value greater than or equal to the target service level SL. target .

[0124] ④ Replenishment frequency constraints:

[0125] The goal is to balance replenishment costs and operational efficiency. Within a certain timeframe, the frequency of replenishment affects replenishment costs; therefore, it's necessary to control the frequency of replenishment while meeting demand. Let the total replenishment cost be C. replenishment Total operating costs C total Including inventory holding costs and replenishment costs, a constraint on the number of replenishments, n, can be introduced to balance replenishment costs and operational efficiency. Assuming that within a time period T, total operating costs are related to the number of replenishments and the safety stock level, the replenishment frequency constraint can be expressed as:

[0126] C total =h×S(t)×T+n×C r ≤C max

[0127] Among them, C max This is the upper limit of total operating costs, and the number of replenishment times n also needs to meet certain practical operational requirements;

[0128] For example, such as:

[0129] n min ≤n≤n max

[0130] Where, n min and n max These are the lower and upper limits for the number of replenishments, ensuring that the number of replenishments is neither too few, leading to an increased risk of stockouts, nor too many, leading to excessive replenishment costs.

[0131] In this embodiment of the invention, the parameters of the genetic algorithm are adjusted in real time using comprehensive performance indicators, as follows:

[0132] P(t)=w1·TC(t)+w2·SL(t)+w3·FR(t)

[0133] Wherein, TC(t) is the total cost, including inventory holding cost, ordering cost, stockout cost, etc.; SL(t) is the service level, reflecting the order fulfillment rate; FR(t) is the replenishment frequency, characterizing the system operating efficiency; w1, w2, and w3 are the corresponding weights, used to balance multiple optimization objectives.

[0134] In one alternative implementation, the weight calculation formula can be:

[0135] Suppose there are n time periods of data, and the total cost, service level, and replenishment frequency for the t-th time period are TC(t), SL(t), and FR(t), respectively. Standardize this data to obtain the standardized matrix X = (x tj ), where x ij Let represent the standardized value of the j-th indicator in the t-th time period. For positive and negative indicators, it is represented as:

[0136]

[0137] Among them, y tj This is the original data, max(y) j ) and minx(y j ) are the maximum and minimum values ​​of the j-th indicator, respectively.

[0138] The weight of the j-th indicator in the t-th time period is calculated as follows:

[0139]

[0140] The entropy value of the j-th index is calculated as follows:

[0141]

[0142] Where, when p tj When = 0, define p tj ln(p tj ) = 0.

[0143] The coefficient of variation for the j-th indicator is expressed as: g j =1-e j

[0144] Calculate the weights, the weight w of the j-th indicator. j for:

[0145]

[0146] In one alternative implementation, a particle swarm optimization algorithm can be used to find the optimal safety stock level. Specifically, the possible range of values ​​for the safety stock level is defined as the search space for the particles, with each particle representing a potential safety stock solution. Particles search for the optimal solution by continuously updating their positions within the search space. In each iteration, the total cost, service level, and replenishment frequency associated with the safety stock level for each particle are calculated, and the fitness value is calculated using a comprehensive performance index formula. By continuously iterating and updating the particle's velocity and position, the particles gradually converge towards the global optimum, ultimately obtaining the optimal safety stock level.

[0147] In another alternative implementation, simulated annealing can be used to find the optimal safety stock level. Specifically, starting from an initial safety stock level solution, a new solution is randomly generated in the neighborhood of the current solution. The difference in comprehensive performance index ΔP between the new solution and the current solution is calculated. If ΔP < 0, the new solution is accepted as the current solution; if ΔP > 0, the new solution is accepted with probability P. As the algorithm progresses, the temperature T is gradually reduced according to a certain cooling strategy. During the cooling process, new solutions are continuously searched. When the temperature drops sufficiently or the stopping condition is met, the current solution is output as the optimal safety stock level.

[0148] It should be noted that this invention constructs an objective function using replenishment lead time, demand standard deviation, and a dynamic safety factor, precisely matching the actual needs of inventory management. The dynamic safety factor adjusts with market changes, making the objective function highly adaptable. Four constraints—inventory holding cost, storage capacity, service level, and replenishment frequency—are applied from the dimensions of cost budget, space constraints, service quality, and operational efficiency, ensuring that the solution meets the actual operational needs of the enterprise and avoiding cost waste or service gaps due to excessively high or low inventory levels. A comprehensive performance index based on total cost, service level, and replenishment frequency can comprehensively evaluate the merits of inventory strategies. By adjusting the genetic algorithm parameters in real time, the algorithm iterates and seeks optimization in multi-objective optimization, ultimately obtaining the optimal safety stock level. This reduces the total costs of inventory holding, ordering, and stockouts while improving order fulfillment rate and system operating efficiency, effectively solving the problem of traditional inventory management's difficulty in balancing multiple objectives.

[0149] Example 3 illustrates an inventory control method based on intelligent algorithms. It should be noted that the technical solution of this inventory control system based on intelligent algorithms belongs to the same concept as the aforementioned inventory control method based on intelligent algorithms. Details not described in detail in this embodiment can be found in the description of the aforementioned inventory control method based on intelligent algorithms.

[0150] This embodiment also provides an inventory control system based on intelligent algorithms, including:

[0151] The data acquisition module is used to collect multi-dimensional data from the warehouse.

[0152] The prediction module is used to input multi-dimensional data into the prediction model, obtain warehouse inventory characteristics through the prediction model, and obtain the predicted value of material consumption based on the inventory characteristics.

[0153] The calculation module is used to calculate the dynamic safety factor of the warehouse based on the predicted value of material consumption and the constraints; based on the dynamic safety factor, an objective function is designed, and the objective function is solved by a genetic algorithm to obtain the optimal safety stock level of the warehouse.

[0154] This embodiment also provides an electronic device suitable for inventory control based on intelligent algorithms, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the inventory control method based on intelligent algorithms as proposed in the above embodiment.

[0155] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the inventory control method based on intelligent algorithms as proposed in the above embodiments.

[0156] The storage medium proposed in this embodiment and the inventory control method based on intelligent algorithms proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0157] Based on the above description of the implementation methods, those skilled in the art will clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0158] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An inventory control method based on intelligent algorithms, characterized in that, include: Collect multi-dimensional data on warehousing; The multi-dimensional data is input into the prediction model, the prediction model is used to obtain the warehouse inventory characteristics, and the predicted value of material consumption is obtained based on the inventory characteristics. Based on the predicted consumption of the materials and the constraints, the dynamic safety factor of the warehouse is calculated. Based on the dynamic safety factor, an objective function is designed, and the objective function is solved using a genetic algorithm to obtain the optimal safety stock level for warehousing.

2. The inventory control method based on intelligent algorithms as described in claim 1, characterized in that, The multi-dimensional data is input into a prediction model, and the prediction model is used to obtain warehouse inventory characteristics. Based on the inventory characteristics, a predicted value of material consumption is obtained, including: Analyze the time series data in the multi-dimensional data to obtain the trend and seasonal characteristics of material demand, and obtain a first predicted value based on the trend and seasonal characteristics; Analyze the external influencing factors in the multi-dimensional data, establish a regression equation, and obtain the second predicted value; The first and second predicted values ​​are weighted and calculated to obtain the predicted value of material consumption.

3. The inventory control method based on intelligent algorithms as described in claim 2, characterized in that, Based on the predicted consumption of the materials and the constraints, the dynamic safety factor of the warehouse is calculated, including: The constraints include the target service level and the volatility of material demand; Based on the predicted value of the material consumption, obtain the prediction mean absolute error; Based on the target service level, the volatility of material demand, and the average absolute error of the forecast, the dynamic safety factor of the warehouse is calculated.

4. The inventory control method based on intelligent algorithms as described in claim 3, characterized in that, Based on the dynamic safety factor, an objective function is designed, and the objective function is solved using a genetic algorithm to obtain the optimal safety stock level for warehousing, including: Based on the replenishment lead time and demand standard deviation of acquired materials, and combined with the aforementioned dynamic safety factor, an objective function is designed. Set up a genetic algorithm to solve the constraints, and then use the genetic algorithm to solve the objective function. Based on total cost, service level, and replenishment frequency, a comprehensive performance index value is calculated, and the genetic algorithm parameters are optimized and adjusted based on the comprehensive performance index value.

5. The inventory control method based on intelligent algorithms as described in claim 4, characterized in that, The constraints that the genetic algorithm solves include inventory holding cost constraints, storage capacity constraints, service level constraints, and replenishment frequency constraints.

6. The inventory control method based on intelligent algorithms as described in claim 4, characterized in that, The objective function is expressed as: Where S(t) is the optimal safety stock level, σ(t) is the standard deviation of demand, k(t) is the dynamic safety factor, and L is the replenishment lead time.

7. The inventory control method based on intelligent algorithms as described in claim 5, characterized in that, The formula for calculating the dynamic safety factor of warehousing is: k(t)=α·CSL+β·MAE(t)+γ·VOL(t) Where k(t) is the dynamic safety factor, CSL is the target service level, MAE(t) is the mean absolute error, VOL(t) is the demand volatility, and α, β, and γ are weighting coefficients.

8. An inventory control system based on intelligent algorithms, employing the method described in any one of claims 1-7, characterized in that, include: The data acquisition module is used to collect multi-dimensional data from the warehouse. The prediction module is used to input the multi-dimensional data into the prediction model, obtain the warehouse inventory characteristics through the prediction model, and obtain the predicted value of material consumption based on the inventory characteristics. The calculation module is used to calculate the dynamic safety factor of the warehouse based on the predicted value of the material consumption and the constraints. Based on the dynamic safety factor, an objective function is designed, and the objective function is solved using a genetic algorithm to obtain the optimal safety stock level for warehousing.

9. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the inventory control method based on intelligent algorithms according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores computer-executable instructions that, when executed by a processor, implement the steps of the inventory control method based on intelligent algorithms as described in any one of claims 1 to 7.

Images