Enterprise operation optimization method, device, equipment and storage medium

Abstract

The present application relates to the field of data processing, and discloses a method, device, equipment and storage medium for optimizing enterprise operations. By acquiring multi-source enterprise data sets, using association rule algorithms to pre-process the multi-source enterprise data sets, an enterprise operation optimization model is established, including an enterprise management layer, a capital optimization layer and a resource scheduling layer; the initial multi-source enterprise data set is input into the enterprise operation optimization model for training, the enterprise situation prediction state is input into the digital twin model for simulation operation, the enterprise operation data is adjusted according to the simulation operation results, the initial enterprise operation data is selected based on the MLP multi-layer perceptron, and the enterprise operation is adjusted in real time according to the target enterprise operation data. Global collaborative optimization from production to sales, from capital to supply chain is achieved. It improves production efficiency and capital flow efficiency, and has greater flexibility and competitiveness in a dynamic market environment.

Description

Technical Field

[0001] The present application relates to the field of intelligent manufacturing technology, and in particular to an enterprise operation optimization method, device, equipment and storage medium. Background Art

[0002] As the transformation and upgrading of the manufacturing industry enters the era of intelligence, traditional production management, supply chain management and capital flow are facing increasingly complex challenges. Modern manufacturing enterprises not only need efficient and flexible production management, but also need intelligent systems that can dynamically respond to market demand, supply chain changes and capital flow. In this process, enterprise operation optimization (Enterprise Situation Generation) has become a key concept. Enterprise operation optimization refers to the real-time analysis and prediction of the current situation and future trends of the enterprise based on multi-dimensional information such as real-time data, market changes, production conditions, supply chain and capital flow, providing intelligent support for decision-makers, and ensuring the efficient allocation and utilization of resources, time, funds and other factors. Therefore, how to establish an efficient enterprise operation optimization model and intelligently adjust production plans, dispatch resources and optimize capital flow according to changes in capital chain and market demand, and enhance the comprehensive competitiveness of enterprises is a technical problem that needs to be solved at this stage. Summary of the invention

[0003] The present application provides a method for optimizing enterprise operations, which can store asset registration in a database and audit asset change request information through a multiple audit mechanism to ensure the security of assets.

[0004] In a first aspect, an embodiment of the present application provides a method for optimizing enterprise operations, the method comprising the following steps:

[0005] Acquire a multi-source enterprise data set, and perform data preprocessing on the multi-source enterprise data set using an Apriori association rule algorithm to obtain an initial multi-source enterprise data set;

[0006] Establishing an enterprise operation optimization model, wherein the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer;

[0007] Inputting the initial multi-source enterprise data set into the enterprise operation optimization model for training to obtain the enterprise situation prediction status;

[0008] Inputting the enterprise situation prediction state into the digital twin model for simulation operation, adjusting the enterprise operation data according to the simulation operation result, and obtaining the initial enterprise operation data;

[0009] The initial enterprise operation data is optimized based on the MLP multi-layer perceptron to obtain target enterprise operation data, and the enterprise operation is adjusted in real time according to the target enterprise operation data.

[0010] In a possible implementation, the step of obtaining a multi-source enterprise data set and preprocessing the multi-source enterprise data set using an Apriori association rule algorithm to obtain an initial multi-source enterprise data set includes:

[0011] Every 5 minutes, the temperature sensor, pressure sensor and speed sensor in the production equipment are used to collect the operation data in the production equipment to obtain the equipment operation data set, and the production order information is extracted from the production management system database to obtain the production order data set;

[0012] Obtain customer order details from the sales order management system to obtain a customer order data set; collect sales channel data and customer feedback data sets, obtain fund income and expenditure details from the financial database to obtain a fund income and expenditure data set; extract cost accounting data and record cost items to obtain a cost accounting data set and a financial statement data set;

[0013] Obtain employee attendance records, performance evaluation data and employee training information through the employee attendance system to obtain employee attendance data sets, performance evaluation data sets and employee training data sets;

[0014] By obtaining market research report data sets, industry dynamic information data sets and macroeconomic indicator data sets from the database, the collected data sets are integrated to obtain multi-source enterprise data sets.

[0015] In a possible implementation, the step of obtaining a multi-source enterprise data set and preprocessing the multi-source enterprise data set using an Apriori association rule algorithm to obtain an initial multi-source enterprise data set further includes:

[0016] Using the mean filling method to fill missing values ​​for the numerical data in the multi-source enterprise data set, and performing mode filling for the categorical data in the multi-source enterprise data set, to obtain a first multi-source enterprise data set;

[0017] For the temperature value T in the first multi-source enterprise data set i Calculate the mean μT and standard deviation σT. The standard deviation calculation formula is:

[0018] It is assumed that if the data with μT±3σT exceeds the threshold range, it is regarded as an outlier and the outlier is deleted to obtain the second multi-source enterprise data set;

[0019] Normalizing the second multi-source enterprise data set to obtain a normalized multi-source enterprise data set;

[0020] Convert the normalized multi-source enterprise data set into a transaction data set Q for association rule mining, and regard each customer order and its related production information, product information, customer information and sales channel information as a transaction;

[0021] Set a minimum support threshold, use the Apriori algorithm to scan the transaction data set Q, count the number of occurrences of each single item, and find a frequent item set L that meets the minimum support; generate a candidate item set C through the frequent item set L;

[0022] Scan the transaction data set Q to calculate the support of the candidate item set C, filter out the frequent item set L2 that meets the minimum support, and repeat the steps until no new frequent item sets are generated to obtain the initial multi-source enterprise data set.

[0023] In a possible implementation, the enterprise operation optimization model is established, and the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer, including:

[0024] The enterprise management layer includes a strategic goal setting model and a decision coordination model, which uses the AHP hierarchical analysis method to set the strategic goal setting model, and establishes a decision coordination model based on the information entropy decision coordination mechanism; establishes an information entropy decision coordination mechanism, and the decision coordination mechanism includes calculating the information entropy of the decision information of the capital optimization layer and the resource scheduling layer;

[0025] Use the analytic hierarchy process (AHP) to determine the key objectives of enterprise operations, which include at least profitability, market competitiveness and sustainable development capabilities, build a target hierarchy, and determine the relative importance weights of each target;

[0026] The fund optimization layer includes a fund optimization configuration model and a fund optimization allocation model, a linear programming model is used as the fund forecasting model, and a linear programming model is used as the fund optimization allocation model;

[0027] The resource scheduling layer includes a production resource scheduling model and a supply chain resource scheduling model. A constraint programming model is used as the production resource scheduling model, and a simulated annealing algorithm model is used as the supply chain resource scheduling model.

[0028] In a possible implementation, the enterprise operation optimization model includes at least an enterprise management layer, a capital optimization layer, and a resource scheduling layer, and also includes:

[0029] Utilize the company's management team to formulate the company's overall operational strategy and goals, and determine annual profit growth targets and market share improvement targets based on market trends and the company's own resource status;

[0030] Based on the capital optimization layer, according to historical sales data, market trends, sales channel expansion, and capital forecasting models, sales revenue for each future time period is predicted;

[0031] Based on the production order demand, equipment capacity and employee skill level at the resource scheduling layer, production tasks are allocated to different production lines, equipment and work shifts; the equipment operating status, raw material inventory level and employee workload during the production process are monitored, and the production scheduling plan is adjusted.

[0032] In a possible implementation, the enterprise situation prediction state is input into the digital twin model for simulation operation, and the enterprise operation data is adjusted according to the simulation operation result to obtain the initial enterprise operation data, including:

[0033] Establishing a digital twin model in which a production system, a supply chain system, and a personnel organization system are virtually mapped;

[0034] The enterprise situation prediction state is input into the digital twin model for simulation operation, and the initial scenario parameters of the simulation operation are determined according to the enterprise situation prediction state;

[0035] Performing simulation operation according to the initial scenario parameters, evaluating the simulation operation results according to the set enterprise operation goals and performance indicators, comparing the indicator values ​​obtained by simulation with the target values ​​expected by the enterprise, and obtaining enterprise operation difference data; the enterprise operation difference data at least includes production efficiency data, cost-effectiveness data, quality indicator data, supply chain performance data and financial indicator data;

[0036] The enterprise situation prediction state is adjusted according to the enterprise operation difference data, and the adjusted enterprise situation prediction state is input into the digital twin model again for simulation operation. When the difference between the data obtained from the operation and the enterprise operation target is less than 1%, the initial enterprise operation data is obtained.

[0037] In a possible implementation, the initial enterprise operation data is selected based on the MLP multilayer perceptron to obtain target enterprise operation data, and the enterprise operation is adjusted in real time according to the target enterprise operation data, including:

[0038] Acquire initial enterprise operation data, and select the initial enterprise operation data using an MLP multilayer perceptron; the MLP multilayer perceptron includes at least an input layer, a plurality of hidden layers, and an output layer;

[0039] After processing the initial enterprise operation data, it is input through the input layer of the MLP multilayer perceptron. For the i-th sample of the k-th batch, the input of the j-th node of the first layer of the hidden layer is:

[0040]

[0041] in, is the weight from the input layer to the first hidden layer, is the bias of the jth node in the first layer of the hidden layer; the output of the jth node in the first layer of the hidden layer is obtained by performing nonlinear transformation through the ReLU activation function:

[0042] In the hidden layer calculation, the input of the jth node in the hidden layer p+1 is: in, is the weight of the hidden layer from layer p to layer p+1, is the bias of the jth node in the p+1th hidden layer, is the output of the i-th node in the p-th layer of the hidden layer. After the nonlinear transformation by the ReLU activation function, the output of the j-th node in the p+1-th layer of the hidden layer is obtained:

[0043] In the calculation of the output layer, the number of output layer nodes is set to n. For the i-th sample in the k-th batch, the output of the j-th node in the output layer is in is the weight from the last hidden layer to the output layer, is the bias of the jth node in the output layer, is the output of the i-th node in the last layer of the hidden layer;

[0044] Using MES mean square error as the loss function, let the true value of the kth batch be Then the loss function is:

[0045] Calculate the gradient of the loss function with respect to the output layer weights and biases, and then use the chain rule to calculate the gradient of the hidden layer weights and biases. Use the Adam optimization algorithm to update the weights and biases to obtain the prediction results for each data combination. Sort the data combinations according to the scores, and select the combination with the highest score as the target enterprise operation data.

[0046] In a second aspect, the present application provides an enterprise operation optimization device, wherein the process card optimization device comprises the following modules:

[0047] A data processing module is used to obtain a multi-source enterprise data set, and perform data preprocessing on the multi-source enterprise data set using an Apriori association rule algorithm to obtain an initial multi-source enterprise data set;

[0048] A model building module, used to build an enterprise operation optimization model, wherein the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer;

[0049] A situation prediction module, used for inputting the initial multi-source enterprise data set into the enterprise operation optimization model for training to obtain the enterprise situation prediction state;

[0050] An operation simulation module is used to input the enterprise situation prediction status into the digital twin model for simulation operation, and adjust the enterprise operation data according to the simulation operation results to obtain initial enterprise operation data;

[0051] The data adjustment module is used to select the initial enterprise operation data based on the MLP multi-layer perceptron to obtain target enterprise operation data, and make real-time adjustments to the enterprise operation according to the target enterprise operation data.

[0052] In a third aspect, the present application provides an enterprise operation optimization device, comprising: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor calls the instructions in the memory so that the enterprise operation optimization device executes each step of the above-mentioned enterprise operation optimization method.

[0053] In a fourth aspect, the present application provides a computer-readable storage medium, in which instructions are stored, and when the computer-readable storage medium is run on a computer, the computer executes each step of the above-mentioned enterprise operation optimization method.

[0054] Based on the method provided by the present application, by obtaining a multi-source enterprise data set, the multi-source enterprise data set is preprocessed using the Apriori association rule algorithm to obtain an initial multi-source enterprise data set; an enterprise operation optimization model is established, and the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer; the initial multi-source enterprise data set is input into the enterprise operation optimization model for training to obtain the enterprise situation prediction state; the enterprise situation prediction state is input into the digital twin model for simulation operation, and the enterprise operation data is adjusted according to the simulation operation results to obtain the initial enterprise operation data; the initial enterprise operation data is selected based on the MLP multi-layer perceptron to obtain the target enterprise operation data, and the enterprise operation is adjusted in real time according to the target enterprise operation data. Global collaborative optimization from production to sales, from capital to supply chain is realized. This technical solution helps enterprises achieve intelligent decision-making in many aspects such as efficient production, resource optimization, and capital flow management through intelligent and adaptive optimization mechanisms, improves production efficiency and capital flow efficiency, and has greater flexibility and competitiveness in a dynamic market environment. Through these innovations, enterprises can cope with complex production environments and changes in market demand to ensure the healthy operation of production and finance. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1This is a flow chart of an embodiment of the first enterprise operation optimization method provided in the embodiments of the present application;

[0056] Figure 2 This is a flow chart of an embodiment of the second enterprise operation optimization method provided in the embodiments of the present application;

[0057] Figure 3 This is a flow chart of an embodiment of the third enterprise operation optimization method provided in the embodiments of the present application;

[0058] Figure 4 A schematic diagram of the structure of the first enterprise operation optimization device provided in the embodiment of the present application;

[0059] Figure 5 A schematic diagram of the structure of an enterprise operation optimization device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0060] The embodiment of the present application provides a method for optimizing enterprise operations, which can store asset registration in a database and audit asset change request information through a multiple audit mechanism, thereby solving the problem of poor security of assets in daily management.

[0061] The technical scheme in the embodiment of the present application will be described below in conjunction with the accompanying drawings in the embodiment of the present application. Wherein, the terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable in appropriate circumstances, so that the embodiments described here can be implemented in a sequence other than the content illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, the process, method, system, product or equipment comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0062] It is understandable that any part of the data acquisition or collection involved in this application has been authorized by the user.

[0063] It is understandable that the executor of the present application may be an enterprise operation optimization device, or a mobile terminal or a server, which is not specifically limited here.

[0064] The specific process of the embodiment of the present application is described below. Figure 1 , Figure 1 The following is a flow chart of an embodiment of a method for optimizing enterprise operations provided in an embodiment of the present application, including:

[0065] 101. Obtain a multi-source enterprise data set, and use the Apriori association rule algorithm to perform data preprocessing on the multi-source enterprise data set to obtain an initial multi-source enterprise data set;

[0066] It can be understood that the operating data in the production equipment is collected every 5 minutes through the temperature sensors, pressure sensors and speed sensors in the production equipment to obtain the equipment operation data set, and the production order information is extracted from the production management system database to obtain the production order data set; the customer order details are obtained from the sales order management system to obtain the customer order data set; the sales channel data and customer feedback data set are collected, and the capital income and expenditure details are obtained from the financial database to obtain the capital income and expenditure data set; the cost accounting data is extracted and the cost items are recorded to obtain the cost accounting data set and the financial statement data set; the employee attendance records, performance evaluation data and employee training information are obtained through the employee attendance system to obtain the employee attendance data set, performance evaluation data set and employee training data set; the market research report data set, industry dynamic information data set and macroeconomic indicator data set are obtained from the database, and the collected data sets are integrated to obtain a multi-source enterprise data set.

[0067] It can be understood that the missing values ​​of the numerical data in the multi-source enterprise data set are filled by the mean filling method, and the mode of the categorical data in the multi-source enterprise data set is filled to obtain the first multi-source enterprise data set;

[0068] For the temperature value T in the first multi-source enterprise data set i Calculate the mean μT and standard deviation σT. The standard deviation calculation formula is:

[0069] It is set that if the data μT±3σT exceeds the threshold range, it is regarded as an outlier, and the outlier is deleted to obtain the second multi-source enterprise data set; the second multi-source enterprise data set is normalized to obtain the normalized multi-source enterprise data set; the normalized multi-source enterprise data set is converted into a transaction data set Q for association rule mining, and each customer order and its related production information, product information, customer information and sales channel information are regarded as a transaction; the minimum support threshold is set, and the Apriori algorithm is used to scan the transaction data set Q, and the number of occurrences of each single item is counted to find the frequent item set L that meets the minimum support; the candidate item set C is generated through the frequent item set L; the transaction data set Q is scanned to calculate the support of the candidate item set C, and the frequent item set L2 that meets the minimum support is screened out, and the steps are repeated until no new frequent item sets are generated to obtain the initial multi-source enterprise data set.

[0070] 102. Establish an enterprise operation optimization model, which at least includes the enterprise management layer, capital optimization layer and resource scheduling layer;

[0071] It can be understood that the enterprise management level includes a strategic goal setting model and a decision coordination model. The AHP hierarchical analysis method is used as the strategic goal setting model, and the decision coordination model is established based on the decision coordination mechanism of information entropy; an information entropy decision coordination mechanism is established, and the decision coordination mechanism includes calculating the information entropy of the decision information of the capital optimization layer and the resource scheduling layer; the hierarchical analysis method AHP is used to determine the key goals of the enterprise operation, and the key goals include at least profitability, market competitiveness and sustainable development capabilities, and a goal hierarchy is constructed to determine the relative importance weights between each goal; the capital optimization layer includes a capital optimization allocation model and a capital optimization allocation model, and a linear programming model is used as a capital forecasting model, and a linear programming model is used as a capital optimization allocation model; the resource scheduling layer includes a production resource scheduling model and a supply chain resource scheduling model, and a constraint programming model is used as a production resource scheduling model, and a simulated annealing algorithm model is used as a supply chain resource scheduling model.

[0072] It is understandable that the enterprise management level is used to formulate the overall operational strategy and goals of the enterprise, determine the annual profit growth target and market share improvement target based on market trends and the enterprise's own resource conditions; the capital optimization layer is based on historical sales data, market trends, sales channel expansion, and the capital forecasting model is used to predict sales revenue in future time periods; the resource scheduling layer is based on production order demand, equipment capacity and employee skill level to allocate production tasks to different production lines, equipment and work shifts; the equipment operating status, raw material inventory level and employee workload in the production process are monitored, and the production scheduling plan is adjusted.

[0073] 103. Input the initial multi-source enterprise data set into the enterprise operation optimization model for training to obtain the enterprise situation prediction status;

[0074] It can be understood that model training: the initial multi-source enterprise data set is divided into a training set and a validation set according to a certain ratio of 70% and 30%. Use the training set to train the enterprise operation optimization model. During the training process, the loss function of the model on the training set (mean square error MSE, cross entropy loss, etc., select the appropriate loss function according to the specific model) is minimized by adjusting the parameters of the model (weights and biases in the neural network, or parameters in the optimization algorithm). During the training process, the model is verified using the validation set, and the performance indicators of the model on the validation set (accuracy, recall rate, F1 value, etc., root mean square error RMSE, etc.) can be used for regression problems to prevent overfitting. When the performance of the model on the validation set no longer improves, stop training.

[0075] Enterprise situation prediction: The pre-processed enterprise data is input into the trained enterprise operation optimization model. The model predicts the enterprise's operation situation in the future based on the input data and the learned pattern, including changes in production capacity, capital flow trends, market sales, etc., to obtain the enterprise situation prediction status.

[0076] 104. Input the enterprise situation prediction status into the digital twin model for simulation operation, adjust the enterprise operation data according to the simulation operation results, and obtain the initial enterprise operation data;

[0077] It is understandable that a digital twin model is established in which the production system, supply chain system, and personnel organization system are virtually mapped;

[0078] Input the enterprise situation prediction status into the digital twin model for simulation operation, and determine the initial scenario parameters of the simulation operation according to the enterprise situation prediction status;

[0079] Conduct simulation operations according to the initial scenario parameters, evaluate the simulation operation results according to the set enterprise operation goals and performance indicators, compare the indicator values ​​obtained by simulation with the target values ​​expected by the enterprise, and obtain enterprise operation difference data; the enterprise operation difference data at least includes production efficiency data, cost-effectiveness data, quality indicator data, supply chain performance data and financial indicator data;

[0080] The enterprise situation forecast status is adjusted according to the enterprise operation difference data, and the adjusted enterprise situation forecast status is input into the digital twin model again for simulation operation. When the difference between the data obtained from the operation and the enterprise operation target is less than 1%, the initial enterprise operation data is obtained.

[0081] 105. Based on the MLP multi-layer perceptron, the initial enterprise operation data is optimized to obtain the target enterprise operation data, and the enterprise operation is adjusted in real time according to the target enterprise operation data.

[0082] It can be understood that the initial enterprise operation data is obtained and the initial enterprise operation data is selected using the MLP multilayer perceptron; the MLP multilayer perceptron includes at least an input layer, a plurality of hidden layers and an output layer;

[0083] After processing the initial enterprise operation data, it is input into the input layer of the MLP multilayer perceptron. For the i-th sample of the k-th batch, the input of the j-th node of the first hidden layer is:

[0084]

[0085] in, is the weight from the input layer to the first hidden layer, is the bias of the jth node in the first layer of the hidden layer; the output of the jth node in the first layer of the hidden layer is obtained by performing nonlinear transformation through the ReLU activation function:

[0086] In the hidden layer calculation, the input of the jth node in the hidden layer p+1 is: in, is the weight of the hidden layer from layer p to layer p+1, is the bias of the jth node in the p+1th hidden layer, is the output of the i-th node in the p-th layer of the hidden layer. After the nonlinear transformation by the ReLU activation function, the output of the j-th node in the p+1-th layer of the hidden layer is obtained:

[0087] In the calculation of the output layer, the number of output layer nodes is set to n. For the i-th sample in the k-th batch, the output of the j-th node in the output layer is in is the weight from the last hidden layer to the output layer, is the bias of the jth node in the output layer, is the output of the i-th node in the last layer of the hidden layer;

[0088] Using MES mean square error as the loss function, let the true value of the kth batch be Then the loss function is:

[0089] Calculate the gradient of the loss function with respect to the output layer weights and biases, and then use the chain rule to calculate the gradient of the hidden layer weights and biases. Use the Adam optimization algorithm to update the weights and biases to obtain the prediction results for each data combination. Sort the data combinations according to the scores, and select the combination with the highest score as the target enterprise operation data.

[0090] Based on the method provided by this application, by obtaining a multi-source enterprise data set, the multi-source enterprise data set is preprocessed using the Apriori association rule algorithm to obtain an initial multi-source enterprise data set; an enterprise operation optimization model is established, and the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer; the initial multi-source enterprise data set is input into the enterprise operation optimization model for training to obtain the enterprise situation prediction state; the enterprise situation prediction state is input into the digital twin model for simulation operation, and the enterprise operation data is adjusted according to the simulation operation results to obtain the initial enterprise operation data; the initial enterprise operation data is selected based on the MLP multi-layer perceptron to obtain the target enterprise operation data, and the enterprise operation is adjusted in real time according to the target enterprise operation data. Global collaborative optimization from production to sales, from capital to supply chain is realized. This technical solution helps enterprises achieve intelligent decision-making in many aspects such as efficient production, resource optimization, and capital flow management through intelligent and adaptive optimization mechanisms, improves production efficiency and capital flow efficiency, and has greater flexibility and competitiveness in a dynamic market environment. Through these innovations, enterprises can cope with complex production environments and changes in market demand to ensure the healthy operation of production and finance.

[0091] See also Figure 2 , Figure 2 The flowchart of the third enterprise operation optimization method provided in the embodiments of the present application includes:

[0092] 201. Obtain a multi-source enterprise data set, and use the Apriori association rule algorithm to perform data preprocessing on the multi-source enterprise data set to obtain an initial multi-source enterprise data set;

[0093] 202. Establish an enterprise operation optimization model, which at least includes the enterprise management layer, capital optimization layer and resource scheduling layer;

[0094] Among them, steps 201-202 are similar to the execution steps of the above-mentioned steps 102-102, and will not be repeated here.

[0095] 203. The enterprise management level includes a strategic goal setting model and a decision coordination model. The AHP hierarchical analysis method is used to set the strategic goal setting model, and the decision coordination model is established based on the decision coordination mechanism of information entropy; the decision coordination mechanism of information entropy is established, and the decision coordination mechanism includes calculating the information entropy of the decision information of the capital optimization layer and the resource scheduling layer;

[0096] Understandably,

[0097] The strategic goal setting model adopts a multi-objective decision analysis method, the analytic hierarchy process (AHP). First, the key goals of the enterprise operation, profitability, market competitiveness, sustainable development capabilities, etc., are determined to build a goal hierarchy. Then, the relative importance weights of each goal are determined through expert evaluation or data analysis.

[0098] The decision coordination model establishes a decision coordination mechanism based on information entropy. The information entropy of the decision information of the capital optimization layer and the resource scheduling layer is calculated to measure the uncertainty of the information. The weight of each layer of decision is determined according to the information entropy. The smaller the information entropy, the higher the certainty of the decision information at that layer and the greater the weight. Then, the decisions of each layer are coordinated by weighted summation, so that the final decision takes into account all factors and reduces conflicts and uncertainties.

[0099] 204. Use the analytic hierarchy process (AHP) to determine the key objectives of enterprise operations. The key objectives include at least profitability, market competitiveness and sustainable development capabilities. Build a hierarchical structure of objectives and determine the relative importance weights of each objective.

[0100] It can be understood that the genetic algorithm model: encodes the production scheduling plan as a chromosome, each chromosome represents the allocation and execution order of a production task. Define the fitness function, and comprehensively evaluate the advantages and disadvantages of chromosomes based on factors such as production cycle, equipment utilization, and delivery date compliance. Through genetic operations such as selection, crossover, and mutation, the chromosome population is continuously evolved until the optimal chromosome that meets the termination condition (reaching the maximum number of iterations or finding the optimal solution) is found, and the corresponding production scheduling plan is the optimal plan.

[0101] 205. The capital optimization layer includes a capital optimization allocation model and a capital optimization configuration model. The linear programming model is used as a capital forecasting model, and the linear programming model is used as a capital optimization configuration model.

[0102] It can be understood that the mixed integer programming model: for the procurement, production, distribution and other links in the supply chain, integer decision variables (purchase batch, production quantity, number of distribution vehicles, etc.) and continuous decision variables (inventory level, transportation time, etc.) are defined. Establish an objective function to minimize the total cost of the supply chain, including procurement cost, production cost, inventory cost, transportation cost, etc. Set constraints, such as material balance constraints, inventory capacity constraints, vehicle capacity constraints, supplier supply capacity constraints, etc. Optimize the allocation of supply chain resources by solving mixed integer programming problems. Simulated annealing algorithm model: regard the supply chain resource scheduling problem as an optimization problem, define the solution space (different combinations of procurement plans, production plans, and distribution plans) and the objective function (total cost or total profit of the supply chain). The simulated annealing algorithm starts from an initial solution, generates new solutions through random perturbations, and accepts new solutions according to the Metropolis criterion (accepting inferior solutions with a certain probability to avoid falling into local optimality). As the number of iterations increases, the probability of accepting inferior solutions is gradually reduced, and finally an approximate optimal solution is found. In the solution process, the solution is reasonably evaluated and adjusted in combination with the actual characteristics and constraints of the supply chain.

[0103] 206. The resource scheduling layer includes a production resource scheduling model and a supply chain resource scheduling model. The constraint programming model is used as the production resource scheduling model, and the simulated annealing algorithm model is used as the supply chain resource scheduling model.

[0104] It is understandable that production resource scheduling reasonably arranges the allocation of production tasks between different production lines, equipment and work shifts based on factors such as production order demand, equipment capacity, and employee skill level. For example, the job scheduling algorithm (first come first served FCFS, shortest job first SJF, priority scheduling, etc.) is used to determine the execution order of production tasks to shorten the production cycle and improve equipment utilization. Real-time monitoring of resource usage, equipment operating status, raw material inventory level, employee workload, etc. in the production process is carried out to adjust the production scheduling plan in a timely manner. For example, when a certain equipment fails, the production task it is executing is reallocated to other available equipment, and the arrangement of subsequent production tasks is adjusted to ensure production continuity. Supply chain resource scheduling optimizes the raw material procurement plan and determines the best procurement time, procurement quantity and supplier selection based on factors such as production plan, inventory level, and supplier delivery period. For example, the inventory management model (economic order quantity model EOQ, reorder point model ROP, etc.) is used to calculate the replenishment point and replenishment batch of raw materials to ensure that the supply of raw materials meets production needs without causing excessive inventory backlogs. Coordinate logistics and distribution resources to optimize the entire logistics process from raw material procurement to finished product delivery to customers. This includes selecting the best mode of transportation (road, rail, air, etc.), transportation route planning, and logistics distribution center site selection. For example, the shortest path algorithm (Dijkstra algorithm) is used to determine the shortest transportation route from the factory to the customer to reduce transportation costs.

[0105] 207. Input the initial multi-source enterprise data set into the enterprise operation optimization model for training to obtain the enterprise situation prediction status;

[0106] 208. Input the enterprise situation prediction status into the digital twin model for simulation operation, adjust the enterprise operation data according to the simulation operation results, and obtain the initial enterprise operation data;

[0107] 209. Based on the MLP multi-layer perceptron, the initial enterprise operation data is optimized to obtain the target enterprise operation data, and the enterprise operation is adjusted in real time according to the target enterprise operation data.

[0108] Among them, steps 207-209 are similar to the execution steps of the above steps 103-105, and will not be repeated here.

[0109] Based on the method provided in the embodiment of the present application, by establishing an enterprise operation optimization model, the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer and a resource scheduling layer. According to the situation report, the system automatically adjusts the production scheduling. When it is predicted that the market demand for a certain product will increase, the production of the product is given priority and the production batches are increased; the inventory management is optimized, the safety stock of best-selling products is increased, and the production plan of slow-selling products is promoted or adjusted; the supply chain scheduling is adjusted, and the supplier is negotiated in advance to increase the supply of raw materials; the capital flow is optimized, and the funds are reasonably arranged for expanding production or R&D investment.

[0110] See also Figure 3 , Figure 3 The flowchart of the third enterprise operation optimization method provided in the embodiments of the present application includes:

[0111] 301. Obtain a multi-source enterprise data set, and use the Apriori association rule algorithm to perform data preprocessing on the multi-source enterprise data set to obtain an initial multi-source enterprise data set;

[0112] 302. Establish an enterprise operation optimization model, which at least includes the enterprise management layer, capital optimization layer and resource scheduling layer;

[0113] 303. Input the initial multi-source enterprise data set into the enterprise operation optimization model for training to obtain the enterprise situation prediction status;

[0114] Among them, steps 301-303 are similar to the execution steps of the above steps 101-103, and the details are not repeated here.

[0115] 304. Establish a digital twin model and virtually map the production system, supply chain system and personnel organization system in the digital twin model;

[0116] It is understandable that the multi-dimensional evaluation method not only focuses on financial indicators, but also comprehensively considers indicators such as operational efficiency, customer satisfaction, and sustainable development. For example, the customer satisfaction survey model (integrating customer feedback data and satisfaction evaluation algorithms in the digital twin model) is used to evaluate changes in customer satisfaction under different operating strategies; from the perspective of sustainable development, indicators such as energy consumption (based on production equipment energy consumption models and energy management data) and waste emissions (combined with production processes and environmental management data) are analyzed to evaluate the impact of corporate operations on the environment and ensure the sustainability of corporate operating strategies in the three dimensions of economy, society, and environment.

[0117] 305. Input the enterprise situation prediction status into the digital twin model for simulation operation, and determine the initial scenario parameters of the simulation operation according to the enterprise situation prediction status;

[0118] It is understandable that the simulation operation results should be comprehensively evaluated according to the set enterprise operation goals and performance indicators. Compare the key indicator values ​​obtained by simulation with the target values ​​expected by the enterprise, and analyze the gaps and deviations. For example, if the enterprise's expected profit target is 1 million yuan per month, but the simulation results show that the profit is only 800,000 yuan, it is necessary to conduct in-depth analysis to see whether it is caused by excessive costs (rising raw material costs, low production efficiency leading to increased unit costs, etc.) or insufficient revenue (market demand does not meet expectations, unreasonable product pricing, etc.); if the target for on-time delivery of orders is 95%, and the simulation result is 90%, it is necessary to check whether there are bottlenecks or delays in various links of the supply chain (procurement cycle, production cycle, logistics distribution time, etc.).

[0119] 306. Conduct simulation operations based on initial scenario parameters, evaluate simulation operation results based on set enterprise operation objectives and performance indicators, compare the indicator values ​​obtained by simulation with the target values ​​expected by the enterprise, and obtain enterprise operation difference data; the enterprise operation difference data at least includes production efficiency data, cost-effectiveness data, quality indicator data, supply chain performance data and financial indicator data;

[0120] It is understandable that the enterprise operation strategy is adjusted and optimized according to the problems and gaps found in the simulation results. In terms of production strategy, if it is found that the low production efficiency is caused by unreasonable equipment configuration or backward production process, it can be considered to update the equipment, optimize the process flow or adjust the production task allocation method; if it is the lack of personnel skills that affects production efficiency, a targeted employee training plan can be formulated. In terms of supply chain strategy, in response to the problem of excessive inventory costs, the inventory management strategy can be adjusted, adopting just-in-time procurement (JIT), optimizing the safety stock level, etc.; for the problem of low on-time delivery of orders, optimize supplier selection and management, strengthen cooperation and communication with suppliers, ensure the timely supply of raw materials, and optimize the logistics distribution network, choose more reliable transportation partners or adjust transportation routes. In terms of financial strategy, according to the results of cost structure analysis, look for opportunities to reduce costs, negotiate with suppliers to reduce the purchase price of raw materials, and optimize the efficiency of capital use (reasonably arrange loan repayment plans, optimize investment portfolios, etc.).

[0121] 307. Adjust the enterprise situation forecast status according to the enterprise operation difference data, and input the adjusted enterprise situation forecast status into the digital twin model again for simulation operation. When the difference between the data obtained from the operation and the enterprise operation target is less than 1%, the initial enterprise operation data is obtained.

[0122] It is understandable that multiple strategy adjustment plans are formulated and re-simulated and evaluated in the digital twin model to compare the effects of different plans. Three cost reduction plans are proposed: Plan A is to optimize the production process to reduce the unit production cost; Plan B is to renegotiate with suppliers to reduce the purchase price of raw materials; Plan C is to reduce inventory backlogs and reduce inventory costs. The business operations of these three plans after implementation are simulated in the digital twin model, and the corresponding cost, profit, production efficiency and other indicators are calculated. The optimal plan is selected or the implementation priority of each plan is determined through comparative analysis. At the same time, in the process of strategy adjustment, the changes in the internal and external environment of the enterprise (market trends, policies and regulations, technological development, etc.) are considered to ensure that the adjusted strategy is forward-looking and adaptable.

[0123] 308. Based on the MLP multi-layer perceptron, the initial enterprise operation data is optimized to obtain the target enterprise operation data, and the enterprise operation is adjusted in real time according to the target enterprise operation data.

[0124] Among them, step 308 is similar to the execution steps of the above-mentioned step 106, and the details are not repeated here.

[0125] Based on the method provided in the embodiment of the present application, the enterprise situation prediction state is input into the digital twin model for simulation operation, and the enterprise operation data is adjusted according to the simulation operation results to obtain the initial enterprise operation data. In the simulation operation process, a series of key operation indicators are monitored and analyzed in real time, including production efficiency (output per unit time, equipment utilization rate, etc.), cost-effectiveness (production cost, inventory cost, transportation cost, total cost and profit, etc.), quality indicators (product qualification rate, defective rate, etc.), supply chain performance (order delivery on time rate, inventory turnover rate, etc.) and financial indicators (capital turnover rate, debt repayment ability index, profitability index, etc.). The changing trends of these indicators are intuitively displayed through data visualization technology (dashboard, charts, etc.), and the relationship between the indicators is deeply analyzed to provide data support for the adjustment of enterprise operation strategies.

[0126] The above describes the enterprise operation optimization method in the embodiment of the present application. The following describes the enterprise operation optimization device in the embodiment of the present application. Figure 4 , Figure 4 A schematic diagram of the structure of an enterprise operation optimization device provided in an embodiment of the present application includes:

[0127] The data processing module 401 is used to obtain a multi-source enterprise data set, and perform data preprocessing on the multi-source enterprise data set using the Apriori association rule algorithm to obtain an initial multi-source enterprise data set;

[0128] Model building module 402, used to build an enterprise operation optimization model, which at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer;

[0129] The situation prediction module 403 is used to input the initial multi-source enterprise data set into the enterprise operation optimization model for training to obtain the enterprise situation prediction state;

[0130] The operation simulation module 404 is used to input the enterprise situation prediction status into the digital twin model for simulation operation, and adjust the enterprise operation data according to the simulation operation results to obtain the initial enterprise operation data;

[0131] The data adjustment module 405 is used to select the initial enterprise operation data based on the MLP multi-layer perceptron to obtain the target enterprise operation data, and make real-time adjustments to the enterprise operation according to the target enterprise operation data.

[0132] Based on the device provided by the embodiment of the present application, by receiving a multi-source enterprise data set, the multi-source enterprise data set is preprocessed using the Apriori association rule algorithm to obtain an initial multi-source enterprise data set; an enterprise operation optimization model is established, and the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer; the initial multi-source enterprise data set is input into the enterprise operation optimization model for training to obtain the enterprise situation prediction state; the enterprise situation prediction state is input into the digital twin model for simulation operation, and the enterprise operation data is adjusted according to the simulation operation results to obtain the initial enterprise operation data; the initial enterprise operation data is selected based on the MLP multi-layer perceptron to obtain the target enterprise operation data, and the enterprise operation is adjusted in real time according to the target enterprise operation data. Global collaborative optimization from production to sales, from capital to supply chain is realized. This technical solution helps enterprises achieve intelligent decision-making in many aspects such as efficient production, resource optimization, and capital flow management through intelligent and adaptive optimization mechanisms, improves production efficiency and capital flow efficiency, and has greater flexibility and competitiveness in a dynamic market environment. Through these innovations, enterprises can cope with complex production environments and changes in market demand to ensure the healthy operation of production and finance.

[0133] See also Figure 5 , Figure 5It is a structural schematic diagram of a process card optimization device for intelligent cloud simulation provided by an embodiment of the present application. The process card optimization device 500 for intelligent cloud simulation may have relatively large differences due to different configurations or performances, and may include one or more processors 510 (for example, one or more processors) and a memory 520, and one or more storage media 530 storing application programs 533 or data 532. Among them, the memory 520 and the storage medium 530 may be short-term storage or persistent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the process card optimization device 500 for intelligent cloud simulation. Furthermore, the processor 510 may be configured to communicate with the storage medium 5630 to execute a series of instruction operations in the storage medium 530 on the process card optimization device 500 for intelligent cloud simulation.

[0134] The process card optimization device 500 of intelligent cloud simulation may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and / or one or more operating systems 531, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It can be understood by those skilled in the art that Figure 5 The structure of the process card optimization equipment for intelligent cloud simulation shown does not constitute a limitation of the process card optimization equipment for intelligent cloud simulation, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0135] The present application also provides an enterprise operation optimization device, wherein the computer device includes a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes each step of the enterprise operation optimization method in the above-mentioned embodiments.

[0136] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes each step of the enterprise operation optimization method.

[0137] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0138] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0139] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0140] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for optimizing enterprise operations, characterized in that: The enterprise operation optimization method comprises the following steps: Acquire a multi-source enterprise data set, and perform data preprocessing on the multi-source enterprise data set using an Apriori association rule algorithm to obtain an initial multi-source enterprise data set; Establishing an enterprise operation optimization model, wherein the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer; Inputting the initial multi-source enterprise data set into the enterprise operation optimization model for training to obtain the enterprise situation prediction status; Inputting the enterprise situation prediction status into the digital twin model for simulation operation, adjusting the enterprise operation data according to the simulation operation results, and obtaining initial enterprise operation data; The initial enterprise operation data is optimized based on the MLP multi-layer perceptron to obtain target enterprise operation data, and the enterprise operation is adjusted in real time according to the target enterprise operation data.

2. The enterprise operation optimization method according to claim 1, characterized in that: The step of obtaining a multi-source enterprise data set and preprocessing the multi-source enterprise data set using an Apriori association rule algorithm to obtain an initial multi-source enterprise data set includes: Every 5 minutes, the temperature sensor, pressure sensor and speed sensor in the production equipment are used to collect the operation data in the production equipment to obtain the equipment operation data set, and the production order information is extracted from the production management system database to obtain the production order data set; Obtain customer order details from the sales order management system to obtain a customer order data set; collect sales channel data and customer feedback data sets, obtain fund income and expenditure details from the financial database to obtain a fund income and expenditure data set; extract cost accounting data and record cost items to obtain a cost accounting data set and a financial statement data set; Obtain employee attendance records, performance evaluation data and employee training information through the employee attendance system to obtain employee attendance data sets, performance evaluation data sets and employee training data sets; By obtaining market research report data sets, industry dynamic information data sets and macroeconomic indicator data sets from the database, the collected data sets are integrated to obtain multi-source enterprise data sets.

3. The enterprise operation optimization method according to claim 1, characterized in that: The step of obtaining a multi-source enterprise data set and preprocessing the multi-source enterprise data set using an Apriori association rule algorithm to obtain an initial multi-source enterprise data set also includes: Using the mean filling method to fill missing values ​​for the numerical data in the multi-source enterprise data set, and performing mode filling for the categorical data in the multi-source enterprise data set, to obtain a first multi-source enterprise data set; For the temperature value T in the first multi-source enterprise data set i Calculate the mean μT and standard deviation σT. The standard deviation calculation formula is: It is assumed that if the data with μT±3σT exceeds the threshold range, it is regarded as an outlier and the outlier is deleted to obtain the second multi-source enterprise data set; Normalizing the second multi-source enterprise data set to obtain a normalized multi-source enterprise data set; Convert the normalized multi-source enterprise data set into a transaction data set Q for association rule mining, and regard each customer order and its related production information, product information, customer information and sales channel information as a transaction; Set a minimum support threshold, use the Apriori algorithm to scan the transaction data set Q, count the number of occurrences of each single item, and find a frequent item set L that meets the minimum support; generate a candidate item set C through the frequent item set L; Scan the transaction data set Q to calculate the support of the candidate item set C, filter out the frequent item set L2 that meets the minimum support, and repeat the steps until no new frequent item sets are generated to obtain the initial multi-source enterprise data set.

4. The enterprise operation optimization method according to claim 1, characterized in that: The enterprise operation optimization model is established, and the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer, including: The enterprise management layer includes a strategic goal setting model and a decision coordination model, which uses the AHP hierarchical analysis method to set the strategic goal setting model, and establishes a decision coordination model based on the information entropy decision coordination mechanism; establishes an information entropy decision coordination mechanism, and the decision coordination mechanism includes calculating the information entropy of the decision information of the capital optimization layer and the resource scheduling layer; Use the analytic hierarchy process (AHP) to determine the key objectives of enterprise operations, which include at least profitability, market competitiveness and sustainable development capabilities, build a target hierarchy, and determine the relative importance weights of each target; The fund optimization layer includes a fund optimization configuration model and a fund optimization allocation model, a linear programming model is used as the fund forecasting model, and a linear programming model is used as the fund optimization allocation model; The resource scheduling layer includes a production resource scheduling model and a supply chain resource scheduling model. A constraint programming model is used as the production resource scheduling model, and a simulated annealing algorithm model is used as the supply chain resource scheduling model.

5. The enterprise operation optimization method according to claim 1, characterized in that: The enterprise operation optimization model is established, and the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer, and also includes: Utilize the company's management team to formulate the company's overall operational strategy and goals, and determine annual profit growth targets and market share improvement targets based on market trends and the company's own resource status; Based on the capital optimization layer, according to historical sales data, market trends, sales channel expansion, and capital forecasting models, sales revenue for each future time period is predicted; Based on the production order demand, equipment capacity and employee skill level at the resource scheduling layer, production tasks are allocated to different production lines, equipment and work shifts; the equipment operating status, raw material inventory level and employee workload during the production process are monitored, and the production scheduling plan is adjusted.

6. The enterprise operation optimization method according to claim 1, characterized in that: The enterprise situation prediction state is input into the digital twin model for simulation operation, and the enterprise operation data is adjusted according to the simulation operation result to obtain the initial enterprise operation data, including: Establishing a digital twin model in which a production system, a supply chain system, and a personnel organization system are virtually mapped; The enterprise situation prediction state is input into the digital twin model for simulation operation, and the initial scenario parameters of the simulation operation are determined according to the enterprise situation prediction state; Performing simulation operation according to the initial scenario parameters, evaluating the simulation operation results according to the set enterprise operation goals and performance indicators, comparing the indicator values ​​obtained by simulation with the target values ​​expected by the enterprise, and obtaining enterprise operation difference data; the enterprise operation difference data at least includes production efficiency data, cost-effectiveness data, quality indicator data, supply chain performance data and financial indicator data; The enterprise situation prediction state is adjusted according to the enterprise operation difference data, and the adjusted enterprise situation prediction state is input into the digital twin model again for simulation operation. When the difference between the data obtained from the operation and the enterprise operation target is less than 1%, the initial enterprise operation data is obtained.

7. The enterprise operation optimization method according to claim 1, characterized in that: The method of selecting the initial enterprise operation data based on the MLP multilayer perceptron to obtain target enterprise operation data, and adjusting the enterprise operation in real time according to the target enterprise operation data, includes: Acquire initial enterprise operation data, and select the initial enterprise operation data using an MLP multilayer perceptron; the MLP multilayer perceptron includes at least an input layer, a plurality of hidden layers, and an output layer; After processing the initial enterprise operation data, it is input through the input layer of the MLP multilayer perceptron. For the i-th sample of the k-th batch, the input of the j-th node of the first layer of the hidden layer is: in, is the weight from the input layer to the first hidden layer, is the bias of the jth node in the first layer of the hidden layer; the output of the jth node in the first layer of the hidden layer is obtained by performing nonlinear transformation through the ReLU activation function: In the hidden layer calculation, the input of the jth node in the hidden layer p+1 is: in, is the weight of the hidden layer from layer p to layer p+1, is the bias of the jth node in the p+1th hidden layer, is the output of the i-th node in the p-th layer of the hidden layer. After the nonlinear transformation by the ReLU activation function, the output of the j-th node in the p+1-th layer of the hidden layer is obtained: In the calculation of the output layer, the number of output layer nodes is set to n. For the i-th sample in the k-th batch, the output of the j-th node in the output layer is in is the weight from the last hidden layer to the output layer, is the bias of the jth node in the output layer, is the output of the i-th node in the last layer of the hidden layer; Using MES mean square error as the loss function, let the true value of the kth batch be Then the loss function is: Calculate the gradient of the loss function with respect to the output layer weights and biases, and then use the chain rule to calculate the gradient of the hidden layer weights and biases. Use the Adam optimization algorithm to update the weights and biases to obtain the prediction results for each data combination. Sort the data combinations according to the scores, and select the combination with the highest score as the target enterprise operation data.

8. An enterprise operation optimization device, characterized in that: The process card optimization device includes the following modules: A data processing module is used to obtain a multi-source enterprise data set, and perform data preprocessing on the multi-source enterprise data set using an Apriori association rule algorithm to obtain an initial multi-source enterprise data set; A model building module, used to build an enterprise operation optimization model, wherein the enterprise operation optimization model at least includes an enterprise management layer, a capital optimization layer, and a resource scheduling layer; A situation prediction module, used for inputting the initial multi-source enterprise data set into the enterprise operation optimization model for training to obtain the enterprise situation prediction state; An operation simulation module is used to input the enterprise situation prediction status into the digital twin model for simulation operation, and adjust the enterprise operation data according to the simulation operation results to obtain initial enterprise operation data; The data adjustment module is used to select the initial enterprise operation data based on the MLP multi-layer perceptron to obtain target enterprise operation data, and make real-time adjustments to the enterprise operation according to the target enterprise operation data.

9. An enterprise operation optimization device, characterized in that: The enterprise operation optimization device comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the enterprise operation optimization device to execute each step of the enterprise operation optimization method as described in any one of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the various steps of the enterprise operation optimization method as described in any one of claims 1-7 are implemented.

Images