Pilot-based cost-reducing and benefit-increasing optimization method

Abstract

The invention relates to the technical field of intelligent optimization and Pilot project process management, in particular to a Pilot-based cost-reducing and benefit-increasing optimization method, which comprises the following steps of: acquiring historical operation data and business constraint conditions of a Pilot project, and preprocessing to construct a sample data set; based on the data set and constraint conditions, constructing an optimization model with cost reduction and benefit increase as targets; solving the optimization model by using an improved particle swarm optimization algorithm to obtain an optimal parameter group meeting constraint conditions, issuing the optimal parameter group to a management and control system for implementation, and collecting implementation effects within 7 days; and evaluating cost reduction and efficiency improvement indexes according to the collected data, if the two indexes reach a set threshold value, determining a popularization scheme, and if the two indexes do not reach the standard, backtracking and correcting a penalty coefficient in the model and solving again until the two indexes reach the standard. According to the method and the device, optimization of parameter configuration is realized, and the popularization efficiency and the resource utilization rate of pilot are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent optimization and Pilot project process management, and particularly relates to a pilot cost reduction and efficiency improvement optimization method. BACKGROUND

[0002] In the practice of new product development, process optimization or process management, enterprises generally use Pilot projects (pilot projects) as an important verification mechanism before change. The Pilot mode verifies the feasibility and profitability of the technical scheme in a limited period and controlled resources through a small-scale and low-risk way, and then guides the subsequent large-scale promotion. With the development of big data analysis, automation control and intelligent optimization technology, system optimization based on Pilot project operation data is gradually becoming a key means to promote enterprise cost reduction and efficiency improvement, especially in manufacturing, energy, construction, environmental protection and other industry scenarios. The relevance between Pilot project parameter setting, resource input and operation results is becoming stronger, and higher requirements are put forward for the structured collection, modeling and optimization analysis of process data.

[0003] CN117236680A discloses a beam factory management optimization system based on particle swarm optimization, which effectively improves production scheduling efficiency and reduces process time and cost through solving and scheduling optimization of beam manufacturing task planning; but this scheme still mainly uses static scheduling and does not involve the operation data recovery and iterative optimization mechanism in the Pilot pilot link; CN110032755A discloses a multi-objective optimization method for urban sewage treatment process under multiple working conditions, designs a multi-objective particle swarm optimization algorithm and introduces a case library to improve the adaptability of the algorithm, but its optimization target is only for water quality and energy consumption, and lacks a clear periodic evaluation and promotion determination mechanism. These existing methods usually have the following shortcomings: the optimization process is executed only once, and a closed-loop feedback is not constructed; the promotion condition is not clear, and lacks quantifiable index determination basis; after optimization failure, the model lacks self-adaptive adjustment capability, and cannot form a model updating mechanism based on data feedback.

[0004] Therefore, the existing Pilot optimization scheme generally has the following technical bottlenecks: it cannot realize the quantitative judgment and iterative feedback mechanism of the operation effect, the optimization process lacks systematicness and stability, and it is difficult to support larger-scale promotion decision; the present application proposes a pilot cost reduction and efficiency improvement optimization method, which constructs an optimization model for cost reduction and efficiency improvement indicators, combines historical data and business constraints, uses an improved particle swarm optimization algorithm to obtain the optimal parameter set, and applies it to the Pilot control system. The system collects the operation data after implementation within a 7-day period, and makes promotion determination according to the double-threshold indicators. When the promotion condition is not met, the penalty coefficient in the model is automatically traced back and corrected for re-optimization, thereby realizing the closed-loop process of complete collection, modeling, solving, implementation, evaluation and re-optimization. SUMMARY

[0005] This section is intended to summarize some aspects of the embodiments of the present application and briefly introduce some preferred embodiments, and some simplifications or omissions can be made in this section as well as in the abstract and title of the present application to avoid obscuring the purpose of this section, the abstract and the title, and such simplifications or omissions cannot be used to limit the scope of the present application.

[0006] In view of the above-mentioned existing problems, the present application is proposed.

[0007] To solve the above technical problems, the present application provides the following technical solutions: collecting historical operation data and business constraint conditions of a pilot project, and preprocessing the historical operation data to obtain a sample data set; An optimization model with the optimal cost reduction and efficiency improvement index as the target is constructed based on the sample data set and the business constraint conditions; An improved particle swarm optimization algorithm is used to solve the optimization model to obtain a parameter group that meets the business constraint conditions; The parameter group is sent to the management and control system of the pilot project in the form of configuration information for operation, the process parameters of the pilot project are adaptively adjusted, and the operation data after implementation is collected within a preset period of 7 days; Based on the operation data after implementation, the cost reduction index and the efficiency improvement index of the parameter group are calculated and compared with the set first threshold and second threshold: When the cost reduction index is greater than or equal to the first threshold, and the efficiency improvement index is greater than or equal to the second threshold, the parameter group is determined as a promotion scheme; When the cost reduction index is less than the first threshold, or the efficiency improvement index is less than the second threshold, the optimization model solving step is returned, the penalty coefficient in the optimization model is corrected, and a new parameter group is re-solved until the promotion scheme is determined.

[0008] As a preferred scheme of the pilot cost reduction and efficiency improvement optimization method, the historical operation data includes the set process parameters, actual operation parameters, unit product energy consumption, unit product cost, equipment operation time, equipment utilization rate, production rhythm, good product rate and number of defective products of the target process in the historical operation stage; The business constraint conditions at least include upper limit of delivery cycle, upper limit of budget, upper limit of product quality consistency, upper limit of equipment load, upper limit of personnel working time, process compliance threshold and emission limit value.

[0009] As a preferred scheme of the pilot cost reduction and efficiency improvement based optimization method, the historical operation data is preprocessed to obtain a sample data set, including: The historical operation data is read, and interpolation algorithm is used to complete the missing items, and three times of standard deviation is executed to remove the abnormal values; The completed numerical data is executed to the minimum and maximum standardization, and all parameter values are linearly mapped to the interval [0, 1]; The data is segmented in a sliding manner with a time window length of 7 days, and each segment forms a record unit; The average process parameter value, average energy consumption value, average output tact, average yield and average equipment utilization rate are calculated for each record unit; Each record unit and its corresponding business constraint condition are combined to form a sample item; All sample items are combined to generate the sample data set.

[0010] As a preferred scheme of the pilot cost reduction and efficiency improvement based optimization method, the sample data set and the business constraint condition are used to construct an optimization model with the optimal cost reduction and efficiency improvement index as the target, including: The decision variable is set to the adjustable process parameter in the sample data set; The first objective function is defined as the minimum unit cost minimization target, and the second objective function is defined as the comprehensive maximum production tact and yield maximization target; The upper limit of the delivery cycle, the upper limit of the budget, the upper limit of the equipment load, the process compliance threshold and the emission limit value recorded in the business constraint condition are used as the model constraint condition; The first objective function and the second objective function are combined into a weighted multi-objective function, and the model constraint condition is embedded; The mapping relationship is established, that is, the input is the process parameter set, and the output is the cost and efficiency index combination; An optimization model including the objective function, the constraint function and the input-output relationship is generated.

[0011] As a preferred scheme of the pilot cost reduction and efficiency improvement based optimization method, the improved particle swarm optimization algorithm is used to solve the optimization model to obtain a parameter group satisfying the business constraint condition, including: The particle swarm size, the maximum iteration number, the inertia weight initial value and the learning factor are set; The parameter vector of each particle is randomly initialized in the parameter feasible region; The fitness value of each particle is calculated, wherein the fitness function is composed of the objective function and the constraint penalty term; The individual optimal solution and the global optimal solution are determined according to the fitness value respectively; updating the parameter vectors of all particles according to the velocity and position updating formula; when the maximum number of iterations is reached, outputting the current global optimal solution as the parameter set.

[0012] As a preferred solution of the pilot cost-reducing and efficiency-increasing optimization method, the parameter set specifically includes: reading all process parameter values in the parameter set; comparing each process parameter value with the corresponding parameter in the current pilot project process database to determine the parameter items that need to be adjusted; determining the target set value, adjustment range, adjustment step, execution sequence, and execution node number for each parameter item that needs to be adjusted; encapsulating the parameter items that need to be adjusted into an optimization scheme file, each parameter item in the file including parameter identification, target value, adjustment range, adjustment step, and execution node number.

[0013] As a preferred solution of the pilot cost-reducing and efficiency-increasing optimization method, the process parameters of the pilot project are adjusted adaptively, including: converting the optimization scheme file into a structured instruction set in the form of key-value pairs; uploading the instruction set to the configuration receiving module of the management and control system through an application interface; parsing the content of the instruction set by the configuration receiving module and generating a control task sequence; modifying the process parameters to the corresponding target values one by one according to the control task sequence by the execution module; and recording the modified parameter state to a running log file.

[0014] As a preferred solution of the pilot cost-reducing and efficiency-increasing optimization method, the post-implementation running data is collected within a preset period of 7 days, including: automatically extracting the running records of each production cycle at the end of each cycle; collecting actual process parameters, unit product cost, total output, yield, average production rhythm, resource utilization rate, total energy consumption, and the number of alarm events; merging the collected data of 7 cycles in chronological order to generate a complete post-implementation running data table; performing standardization processing on the post-implementation running data table consistent with the sample data set, and outputting a standardized running data set.

[0015] As a preferred scheme of the pilot-based cost reduction and efficiency improvement optimization method of the present application, based on the operation data after implementation, the cost reduction index and efficiency improvement index of the parameter set are calculated, including: The unit cost data before and after implementation is read from the standardized operation data set, and the cost reduction index C is calculated, and the calculation formula is: Among them, The unit product cost before implementation, The unit product cost after implementation; The average production rhythm, yield rate and resource utilization rate data before and after implementation are read from the standardized operation data set, and the efficiency improvement index E is calculated, and the calculation formula is: Among them, And The average production rhythm before and after implementation, And The yield rate before and after implementation, And The resource utilization rate before and after implementation, The weight coefficients of the average production rhythm, yield rate and resource utilization rate are respectively; The cost reduction index C and the efficiency improvement index E are output as the input results of the threshold comparison.

[0016] As a preferred scheme of the pilot-based cost reduction and efficiency improvement optimization method of the present application, the setting method of the first threshold and the second threshold, including: Read the cost reduction index and efficiency improvement index statistical data of at least 10 pilot projects of the same type from the historical project database; Calculate the average value of the historical cost reduction index and multiply it by the correction coefficient 1.05 to obtain the first threshold; Calculate the upper quartile of the historical efficiency improvement index and multiply it by the correction coefficient 1 to obtain the second threshold; The first threshold and the second threshold are recorded in the threshold configuration table respectively as the fixed input of the threshold comparison module.

[0017] The present application has the following advantages: the present application not only realizes the conversion from empirical data to optimization model, but also introduces improved swarm intelligence algorithm and quantitative judgment mechanism to ensure the efficiency, controllability and promotion value of the optimization result; at the same time, the self-adaptive ability of the system in the actual working condition fluctuation is further enhanced through the penalty coefficient callback, and the stability and cost benefit of the parameter configuration scheme in the actual landing process are significantly improved. ​​Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. 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. Wherein: Figure 1 This is a flowchart illustrating the cost reduction and efficiency improvement optimization method based on pilot as shown in this invention. Detailed Implementation

[0019] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort should fall within the scope of protection of this invention.

[0021] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0022] According to an embodiment of the present invention, in combination Figure 1 The flowchart shown illustrates a cost reduction and efficiency improvement method based on pilot technology, which specifically includes the following steps: S1. Collect historical operational data and business constraints of the pilot project, and preprocess the historical operational data to obtain a sample dataset. Note the following in this step: The historical operation data files of the pilot project are automatically read from the enterprise's production execution system. To ensure the integrity and time consistency of the data, database query commands are used for batch export. The data fields include: set process parameters (such as temperature, pressure, speed, and flow rate), actual operating parameters (actual values ​​collected in real time by sensors), unit product energy consumption and unit product cost, equipment running time and equipment utilization rate, production cycle time, yield rate and number of defective products. Then read the business constraint condition file from the project management system, read each constraint index in the project configuration table through the configuration interface, and extract parameters including: upper limit of delivery period (such as 10 days), upper limit of budget (such as 50,000 yuan), upper limit of product quality consistency (such as 5%), upper limit of equipment load (such as 85%), upper limit of personnel working hours (such as 40 hours), process compliance threshold (such as 95%), and emission limit value (such as 3.5 kg CO2 / piece); The above two types of data are encoded and standardized according to a unified field format to form a structured input data set, and the index key of each record is uniquely identified by a timestamp and a process number. The storage structure adopts a two-dimensional table form. The structured input data set is used as the input of subsequent data preprocessing, and provides a unified data interface for model training and optimization.

[0023] It should be noted that by establishing a data docking interface between the production execution system and the project management system, the automatic extraction and synchronous formatting of running data and constraint conditions are realized, avoiding the delay and errors of manual sorting, ensuring the integrity and consistency of the input data for subsequent optimization modeling, and realizing the standardized fusion of data standards, significantly improving the data applicability and calculation accuracy in the model construction stage.

[0024] In the preferred embodiment, the extracted historical running data is preprocessed, specifically including the following steps: For data with missing items, linear interpolation algorithm is used for numerical completion; for example, if the observation values of two consecutive time points , are , , and the middle time is missing, then For abnormal values, the three-sigma method is used to remove, that is, if a parameter value satisfies , it is judged as an abnormal item and removed, where is the mean value, is the standard deviation; For all numerical parameters, minimum and maximum normalization processing is performed to linearly map the data to the [0, 1] interval. This normalization process eliminates the scale difference between different dimensions, ensuring the comparability of subsequent multi-objective optimization calculations; Then segment the data in a sliding manner with a time window length of 7 days. The continuous data in each time period is divided into a record unit to reflect the short-term running characteristics. For each record unit, the following average value features are calculated: wherein n represents the number of data entries within the current time window, i represents the i-th data record, represents the process parameter value of the i-th record, represents the average value of all process parameter values within the current time window, represents the energy consumption per product of the i-th record, represents the average energy consumption level within the current time window, represents the production tact of the i-th record, represents the average production tact within the current time window, represents the yield of the i-th record, represents the average yield within the current time window, represents the equipment utilization of the i-th record, represents the average equipment utilization within the current time window.

[0025] For example, if the pilot project runs for 21 days from March 1, 2024 to March 21, 2024, and the window is 7 days, three record units are formed: Record unit 1 (March 1~March 7); Record unit 2 (March 8~March 14); Record unit 3 (March 15~March 21); Each record stores the average process characteristics and energy consumption, tact, yield, equipment utilization, etc. statistical indicators of the corresponding period.

[0026] It should be noted that the preprocessing method combining interpolation completion and abnormality rejection ensures the continuity and reliability of the samples, the normalization processing enhances the comparability of different parameters, and the sliding window segmentation makes the sample data have time representativeness and dynamic sensitivity, which is beneficial to the subsequent optimization model to capture the running trend change and improve the response ability of the optimization model to periodic fluctuations.

[0027] In the preferred embodiment, each record unit is further combined with the business constraint condition in the corresponding time period to form an independent sample entry, and the combination method is: the feature vector of the record unit is spliced with the business constraint vector to form a composite sample vector: wherein represents the vector splicing operation, and the sample entry s stores the mapping relationship between the process characteristics and the constraint parameters in the time period; Batching all the sample entries formed in all time periods to finally generate a sample dataset wherein is the number of sample entries, the dataset is stored in CSV format and used as the input for optimization model training and validation.

[0028] Preferably, the sample entry combination mechanism integrates process variables and management constraints to form multi-dimensional samples reflecting the coupling relationship between production efficiency, energy consumption cost and process constraints, and through a unified structure storage format, facilitates subsequent algorithm model fast indexing, matrix input and objective function solving, and improves the overall efficiency of data management and optimization calculation.

[0029] S2, based on the sample dataset and business constraint conditions, an optimization model with the optimal cost-reducing and efficiency-increasing index as the target is constructed. It needs to be explained that: The decision variables (i.e. parameters that can be adjusted to affect the running cost and production efficiency of the pilot project) are the adjustable process parameters in the sample dataset; It needs to be explained that the value range of each decision variable is jointly limited by historical running data and business constraint conditions, for example, the upper limit of temperature is determined by equipment safety constraints, and the lower limit of flow is limited by emission standards. The decision variables constitute the input vector of the optimization model, and all variables are mapped to the interval [0, 1] after normalization processing. By screening and defining decision variables from the sample dataset, a controllable search area can be realized in the multi-dimensional parameter space, ensuring that the subsequent algorithm solving is performed under the premise of meeting the process boundary conditions, and improving the stability and solving efficiency of the model calculation; The first objective function is defined as the minimum unit cost minimization target, and the second objective function is defined as the comprehensive maximization target of production rhythm and good product rate; Exemplarily, the mathematical expression formula of the first objective function is: Exemplarily, the mathematical expression formula of the second objective function is: The upper limit of the delivery period, the upper limit of the budget, the upper limit of the equipment load, the process compliance threshold and the emission limit value recorded in the business constraint conditions are used as the model constraint conditions, and the constraint function is defined as: The first objective function and the second objective function are fused into a weighted multi-objective function, and the model constraint conditions are embedded. The mathematical expression formula of the weighted multi-objective function is: A mapping relationship is established: the input is a set of process parameters, and the output is a combination of cost and performance indicators; wherein Y is an output indicator vector, including cost reduction indicator C and efficiency improvement indicator E; An optimization model including objective function, constraint function and input-output relationship is generated, and its mathematical expression formula is: wherein, is an adjustable process parameter vector, is a first objective function, is a second objective function, , , respectively represent raw material, energy consumption and labor cost, is output per unit time, is yield rate, is production rhythm, is energy consumption ratio, and α, β, γ are efficiency function weights, and λ is energy consumption penalty coefficient, , , , , respectively are delivery cycle, budget, equipment load, process compliance degree and emission function, , , , , is a corresponding constraint upper limit, , is a multi-objective weighting coefficient, , , , respectively are the upper and lower bounds of the corresponding objective function in the sample data set, is a constraint penalty function, is a penalty factor.

[0030] It should be noted that the comprehensive objective function after normalization, its value range is [0, 1], wherein, when the value tends to 0, it means that the system realizes the minimum cost and the maximum efficiency under the condition of meeting all constraints, representing the highest degree of optimization completion; when the value tends to 1, it means that there is a larger deviation in the objective function, and the decision variable or the penalty coefficient needs to be further adjusted.

[0031] Specifically, within the feasible solution space, the nonlinear mapping of input X and output Y is approximated by using sample data, and the optimal parameter combination is obtained by solving the minimum value of Since the model contains the penalty function term , the search direction can be automatically corrected when the constraint is violated, and convergence within the feasible region is achieved.

[0032] Preferably, this step realizes a multi-objective optimization modeling method of fusing the Pilot operation data with the business constraints, so that the model can not only ensure the realization of the cost reduction target, but also improve the stability of the efficiency improvement index, thereby providing a clear mathematical solution basis for the subsequent improved particle swarm optimization algorithm.

[0033] S3, solving the optimization model by using the improved particle swarm optimization algorithm to obtain a parameter group meeting the business constraint conditions. It should be noted that in this step that: In the preferred embodiment, in order to efficiently solve the optimization model constructed in step S2, the basic parameters of the particle swarm algorithm need to be set, and the population is initialized. Specifically, the system determines the particle dimension according to the number of process parameters to be optimized, and determines the upper and lower limits of the value of each parameter according to the historical operation data and the process boundary conditions, and sets the number of particle swarms (such as 40), the maximum number of iterations (such as 150), the value of the inertia weight range (such as 0.4~0.9) and the learning factor (such as 1.9) according to the project scale and the computing capacity; it is not difficult to understand that the inertia weight adopts a linear decreasing manner in the algorithm running process, so as to enhance the global search ability in the early stage and improve the local convergence precision in the later stage; After completing the parameter setting, the initial particle swarm is randomly generated in the value range allowed by each process parameter in a Latin hypercube sampling manner, so that the initial distribution of all particles uniformly covers the feasible solution space, thereby avoiding the problems of too dense distribution or bias to the boundary in the initial solution set; each particle contains a parameter vector and a velocity vector, and the initial velocity is randomly generated in the safe range, which is used to control the parameter adjustment step, and after the above setting, the initial state of the algorithm is established, thereby laying a stable foundation for subsequent iteration optimization; In each iteration process, the fitness value of all particles needs to be calculated to evaluate their advantages and disadvantages. The fitness function is composed of the objective function and the constraint penalty term, and its mathematical expression is as follows: Among them, represents the fitness value of the particle under the current iteration, is the comprehensive objective function value, is the calculation result of the kth constraint condition, which takes a positive value when the constraint is violated, and K is the total number of constraints, is a dynamic penalty coefficient; In the fitness evaluation process, the following two cases are included: (1) particles meeting the constraints When all the constraint functions of a particle are , it means that it meets all the constraint conditions. At this time, the fitness value is determined by the comprehensive objective function Direct decision, that is, the cost reduction index and the efficiency index of the particle are brought into the objective function, a weighted sum is obtained through the set weight, the performance score corresponding to the particle is obtained, the lower the score, the smaller the cost and the higher the performance, the particle has a better ranking in the ranking; for example, in a certain pilot project, when the unit cost is 92 yuan / piece and the comprehensive performance score is 0.88, the fitness value obtained after weighting through the objective function is 0.12, then the particle is considered to be better than the particle with a fitness value of 0.25; (2) Particle with constraint violation When any constraint function , the corresponding penalty term is automatically superimposed; the specific method is: the over-limit value of each violated constraint term is calculated, and the square of the over-limit value is multiplied by the dynamic penalty coefficient of the constraint , all penalty terms are added to form a total penalty value, and the total penalty value is added to the fitness value; for example, when the delivery period exceeds the allowed upper limit by 3% and the budget exceeds the upper limit by 2%, the over-limit part is squared and weighted to superimpose, so that the fitness value of the particle is significantly increased, and the penalty coefficient increases with the number of iterations to enhance the coerciveness of constraint convergence; in this way, the penalized particle is at a disadvantage in the ranking, and will be guided to move to the feasible solution region in the next iteration; Through this mechanism, the algorithm always maintains the search principle of giving priority to feasible solutions in the entire optimization process, that is, particles that meet the constraints and have good target performance are preferentially selected into the next generation population to realize continuous approximation to the feasible solution space; Through the above optimization design, the particle swarm algorithm in the embodiment can realize efficient solution in the multi-constraint and multi-objective environment of the pilot project, the parameter setting is reasonable and the initialization distribution is balanced, so that the algorithm has good global exploration ability, the fitness function design is combined with the dynamic penalty mechanism, and the compliance with the constraint condition and the rapid convergence of the objective function are ensured; After calculating the fitness value of each particle, the historical optimal solution (i.e. individual optimal) of each particle and the global optimal solution of the current population are automatically updated; specifically, the current fitness value of each particle is compared with the historical optimal fitness value, if the current value is better, the current position of the particle is recorded as the new individual optimal position; at the same time, the particle with the optimal fitness value is selected from all particles as the global optimal solution of the current iteration, if there are multiple particles with the same fitness value, the particle with lower cost and smaller energy consumption is preferentially selected as the global optimal solution to meet the business orientation of cost reduction and efficiency improvement; Through the double-layer update mechanism, the algorithm forms a dynamic balance between local and global search, the individual optimal solution ensures that the particle continuously explores in the direction of its historical optimal solution, and the global optimal solution guides the overall group to move towards the global optimal region, thereby effectively preventing the search from falling into a local minimum value; In each iteration, the velocity and position of the particle need to be updated, which is updated by the following formula: wherein, is the current velocity of the particle, is the current position of the particle, is the historical individual optimal position of the particle, is the global optimal position of the population, is the inertia weight, which decreases with the iteration number, and are the individual learning factor and the population learning factor, respectively, and are random numbers between 0 and 1, which are used to enhance the randomness of the search, is the contraction factor, which is used to prevent the solution from being unstable due to too large velocity; It should be noted that, in order to prevent the particle from exceeding the feasible region of the parameters, when the particle position exceeds the boundary, the system will use the mirror repair strategy to map the out-of-bound position back to within the boundary; at the same time, if the particle has not improved for several generations, the system will implement a small perturbation to some particles with poor fitness to increase the population diversity and avoid premature convergence; through the above mechanism, the algorithm balances the global optimization and local refinement while ensuring the feasibility; When the iteration process meets the termination condition, the final global optimal solution is output as the optimal parameter group, and then all the process parameter values in the parameter group are read and compared with the existing parameters in the process database of the pilot project one by one to find out the parameters that need to be adjusted. For each parameter that needs to be adjusted, the system determines its target set value, allowed adjustment range, adjustment step, execution order, and execution node number; For example, when the optimal parameter group shows that the reaction temperature needs to be adjusted from 175 ℃ to 182 ℃, the system will generate the target value 182 ℃ of this parameter, the allowed interval 170 ℃ to 190 ℃, the single adjustment step 1 ℃, and arrange it as the first step according to the process logic; if the flow parameter needs to be adjusted from 50 L / min to 52 L / min, the step size is set to 0.5 L / min and arranged to be executed in the second step; for example, when the device speed parameter is increased from 840 rpm to 860 rpm, it is arranged to be executed in the third step; The system encapsulates the above parameter items and corresponding information into an optimization scheme file, each parameter contains parameter name, target value, adjustment range, adjustment step, execution order, and node number, etc. The file is stored in a structured format and can be directly recognized and executed by the management and control system, realizing the automatic optimization management of the pilot project process parameters.

[0034] Preferably, the particle swarm optimization solving process of the present application can realize global optimal search of process parameters under the premise of meeting complex business constraints, wherein through control of dynamic weight and shrinkage factor, global exploration in the early stage and local refinement in the later stage are realized; through dynamic penalty mechanism and feasible priority rule, stable convergence of the optimization process to the feasible solution domain is ensured; through boundary repair and disturbance mechanism, the algorithm is prevented from falling into local optimum; therefore, not only the solving efficiency and stability of cost-reducing and benefit-increasing scheme are significantly improved, but also the implementation feasibility of the pilot project in the actual production environment is ensured.

[0035] S4, send the optimal parameter set in the form of configuration information to the management and control system of the pilot project for operation, adaptively adjust the process parameters of the pilot project, and collect the running data after implementation in a preset period of 7 days. It needs to be explained that in this step: convert the optimization scheme file into a structured instruction set in the form of key-value pair structure; upload the instruction set to the configuration receiving module of the management and control system through the application interface; parse the content of the instruction set by the configuration receiving module and generate a control task sequence; modify the process parameters to the corresponding target values item by item by the execution module according to the control task sequence; record the modified parameter state to the running log file.

[0036] Further, the running data after implementation specifically includes: automatically extract the running record of each production cycle at the end of the cycle; collect actual process parameters, unit product cost, total output, yield, average production rhythm, resource utilization rate, total energy consumption, and the number of alarm events; merge the collected data of 7 cycles in chronological order to generate a complete post-implementation running data table; perform standardization processing on the post-implementation running data table consistent with the sample data set, and output the standardized running data set.

[0037] It needs to be explained that in this embodiment, the optimal parameter set is smoothly mapped to the device executable parameter adjustment action, providing auditable and traceable data basis for subsequent threshold judgment, ensuring the consistency and comparability of data before and after implementation, thereby reliably supporting the threshold judgment of the next step; Preferably, compared with the manual parameter adjustment or the extensive mode of one-time centralized issuance in the prior art, the step has the advantages of being replicable, rollbackable and low-risk: firstly, it significantly reduces the safety and compliance risks in the parameter landing process; secondly, it ensures stable data quality within a 7-day observation window and improves the statistical reliability of cost reduction and efficiency improvement evaluation; and thirdly, it provides high-quality backhaul data and detailed operation trajectory for further optimization, thereby improving the success rate of promotion.

[0038] S5, based on the running data after implementation, calculating the cost reduction index and efficiency improvement index of the preferred optimization scheme, and comparing them with the set first threshold and second threshold. It should be noted that the step is: reading the unit cost data before and after implementation from the standardized running data set, calculating the cost reduction index C, and the calculation formula is: wherein, is the unit product cost before implementation, is the unit product cost after implementation; reading the average production rhythm, yield rate and resource utilization rate data before and after implementation from the standardized running data set, calculating the efficiency improvement index E, and the calculation formula is: wherein, and are the average production rhythm before and after implementation, and are the yield rate before and after implementation, and are the resource utilization rate before and after implementation, , , are the weight coefficients of the average production rhythm, yield rate and resource utilization rate, respectively; outputting the cost reduction index C and the efficiency improvement index E as the input results for threshold comparison; when the cost reduction index is greater than or equal to the first threshold, and the efficiency improvement index is greater than or equal to the second threshold, determining the preferred optimization scheme as the promotion scheme; when the cost reduction index is less than the first threshold, or the efficiency improvement index is less than the second threshold, returning to the optimization model solving step, modifying the penalty coefficient in the optimization model, and re-solving to obtain a new optimal parameter group until the promotion scheme is determined.

[0039] Further, the setting method of the first threshold and the second threshold includes: reading at least 10 statistical data of cost reduction index and efficiency improvement index of pilot projects of the same type from the historical project database; calculating the average value of the historical cost reduction index and multiplying it by a correction coefficient 1.05 to obtain the first threshold; The upper quartile of the historical efficiency index is calculated and multiplied by a correction coefficient 1 to obtain a second threshold value; The first threshold value and the second threshold value are recorded in the threshold configuration table as fixed inputs of the threshold comparison module.

[0040] In the preferred embodiment, when the cost reduction index or the efficiency index does not reach the preset threshold value, the system will enter the correction process of the optimization model penalty coefficient to guide the algorithm to more effectively balance the objective function and the constraint condition in the next iteration, and to realize the adaptive adjustment of the optimization direction, wherein: From the last round of solving results, the execution and deviation data of the constraint function are extracted, including the frequency of constraint violation, the overrun amplitude and the duration signal, and according to the constraint type, it is divided into three categories: safety compliance category, economic performance category and stability category. The safety compliance category includes hard constraints such as emission upper limit, equipment load, and personnel work hour upper limit. The economic performance category includes budget, energy consumption, and delivery period, and other cost-related conditions. The stability category includes process compliance, quality consistency, and production line balance conditions. For each category of constraints, the system adjusts the penalty weight according to its violation degree and importance. When the safety compliance category constraint is violated, directly increase its penalty coefficient to a high level, so that the algorithm forces to meet the constraint condition in the next round of solving. For economic performance category constraints, the penalty coefficient is graded and adjusted according to the actual deviation proportion: the smaller the deviation amplitude, the slower the increase, and the larger the deviation amplitude, the faster the increase. The penalty adjustment of the stability category constraint is based on volatility. If the same type of unstable behavior is detected in two consecutive solving, the corresponding penalty value is doubled. The historical penalty trend of each constraint is recorded using a rolling window, and an inertia adjustment mechanism is formed by a trend weighting method. When a certain constraint remains stable and has no violation record in multiple iterations, the system automatically attenuates its penalty strength to expand the search space and restore the flexibility of the model. If the out-of-bound behavior frequently occurs in multiple rounds, the penalty strength of the constraint is continuously enhanced until the preset upper limit value is reached. Through this mechanism, the model can dynamically tighten or relax the constraint degree according to the historical performance to improve the stability of convergence. In the penalty correction, the system adopts a risk stratification strategy. For the safety compliance category constraint, the penalty weight adjustment amplitude is the highest and takes effect immediately. For the economic performance category constraint, the adjustment amplitude is moderate, and a verification solving is required before taking effect to confirm the correct direction. For the stability category constraint, the penalty coefficient is fine-tuned through a secondary adjustment mechanism to prevent convergence stagnation caused by excessive tightening. All correction rules have priority ordering: safety compliance > economic performance > stability. When different types of constraints conflict, the safety category is given priority, and other constraints are in descending order. After each penalty coefficient correction, the system generates a penalty correction record table, recording the following contents: the adjusted constraint term, the penalty value before and after correction, the adjustment range, the trigger reason, the impact round, and the effective range; all records are distinguished by version number and automatically archived in the model version library, ensuring traceability for subsequent solutions. If the subsequent solution shows that the penalty value is too high, resulting in excessive shrinkage of the feasible region, the system will automatically reduce the penalty coefficient according to the decay rule to restore the model activity. After the correction is completed, the system writes the updated penalty coefficient into the optimization model configuration file and triggers a new round of solving task, using the global optimal parameters of the previous round as the warm start point and recalculating the objective function and constraint balance ratio to explore better feasible solutions while maintaining continuity.

[0041] Through the above implementation, the correction of the penalty coefficient realizes intelligent adaptive adjustment of the constraint level, enabling the optimization process to have dynamic evolution capabilities such as self-learning, self-constraint, and self-recovery. This method can automatically adjust the search weight according to historical operation and constraint violation, effectively avoiding constraint failure caused by excessive relaxation and solution stagnation caused by excessive tightening.

[0042] Further, the correction process is combined with the model versioning mechanism, making each optimization adjustment traceable to ensure safety and controllability. Through this mechanism, the algorithm gradually approaches the balanced solution that considers both cost optimization and performance optimization in multiple iterations, thereby achieving long-term stability in cost reduction and efficiency improvement.

[0043] The aforementioned preprocessing method for historical operation data can be performed using existing technical means, which will not be described in detail in this example.

[0044] It should be noted that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and all should be included in the scope of the claims of the present application.

Claims

1. An optimization method for cost reduction and efficiency improvement based on pilot testing, characterized in that, include: Collect historical operational data and business constraints of the pilot project, and preprocess the historical operational data to obtain a sample dataset; An optimization model is constructed based on the sample dataset and the business constraints, with the goal of optimizing cost reduction and efficiency improvement indicators. The optimization model is solved using an improved particle swarm optimization algorithm to obtain a parameter set that satisfies the business constraints. The parameter set is sent to the pilot project's management system in the form of configuration information for operation, adaptively adjusting the pilot project's process parameters, and collecting post-implementation operation data within a preset period of 7 days. Based on the operational data after implementation, the cost reduction and efficiency improvement indicators of the parameter group are calculated and compared with the set first and second thresholds: When the cost reduction indicator is greater than or equal to the first threshold and the efficiency improvement indicator is greater than or equal to the second threshold, the parameter group is determined as the promotion plan; When the cost reduction index is less than the first threshold, or the efficiency improvement index is less than the second threshold, the process returns to the optimization model solution step. By correcting the penalty coefficient in the optimization model, the solution is re-solved to obtain a new set of parameters, until the promotion plan is determined.

2. The optimization method for cost reduction and efficiency improvement based on pilot according to claim 1, characterized in that, The historical operating data includes the set process parameters, actual operating parameters, unit product energy consumption, unit product cost, equipment operating time, equipment utilization rate, production cycle time, yield rate, and number of defective products for the target process during the historical operating phase. The business constraints include at least the upper limit of delivery cycle, upper limit of budget, upper limit of product quality consistency, upper limit of equipment load, upper limit of personnel working hours, process compliance threshold and emission limit.

3. The optimization method for cost reduction and efficiency improvement based on pilot according to claim 1 or 2, characterized in that, The historical operational data is preprocessed to obtain a sample dataset, including: Read the historical operation data, use an interpolation algorithm to fill in missing items, and remove outliers by three times the standard deviation; The completed numerical data is subjected to minimum and maximum standardization, which linearly maps all parameter values ​​to the interval [0,1]. The data is segmented using a sliding method with a time window length of 7 days, and each segment forms a record unit; For each record unit, calculate the average process parameter value, average energy consumption value, average output cycle time, average yield rate, and average equipment utilization rate; Each record unit and its corresponding business constraints are combined to form a sample entry; All sample entries are merged to generate the sample dataset.

4. The optimization method for cost reduction and efficiency improvement based on pilot according to claim 3, characterized in that, Based on the sample dataset and the business constraints, an optimization model is constructed with the goal of optimizing cost reduction and efficiency improvement indicators, including: The decision variables are set as adjustable process parameters in the sample dataset; The first objective function is defined as minimizing unit cost, and the second objective function is defined as maximizing both production cycle time and yield rate. The upper limit of delivery cycle, upper limit of budget, upper limit of equipment load, process compliance threshold and emission limit recorded in the business constraints are used as model constraints. The first objective function and the second objective function are combined into a weighted multi-objective function, and the model constraints are embedded therein; Establish a mapping relationship, where the input is a set of process parameters and the output is a combination of cost and efficiency indicators; Generate an optimization model that includes the objective function, constraint functions, and input-output relationships.

5. The optimization method for cost reduction and efficiency improvement based on pilot according to claim 4, characterized in that, The optimization model is solved using an improved particle swarm optimization algorithm to obtain a parameter set that satisfies the business constraints, including: Set the particle swarm size, maximum number of iterations, initial value of inertia weight, and learning factor; Randomly initialize the parameter vector for each particle within the parameter feasible region; Calculate the fitness value for each particle, where the fitness function is a weighted sum of the objective function and constraint penalty terms; The individual optimal solution and the global optimal solution are determined based on the fitness values. Update the parameter vectors of all particles according to the velocity and position update formula; When the maximum number of iterations is reached, the current global optimal solution is output as the parameter set.

6. The optimization method for cost reduction and efficiency improvement based on pilot according to claim 5, characterized in that, The parameter set specifically includes: Read all process parameter values ​​from the parameter group; Compare the values ​​of each process parameter with the corresponding parameters in the current pilot project process database to determine the parameter items that need to be adjusted; For each parameter that needs adjustment, determine the target setting value, adjustment range, adjustment step size, execution order, and execution node number; The parameters that need to be adjusted are encapsulated into an optimization plan file. Each parameter in the file contains a parameter identifier, target value, adjustment range, adjustment step size, and execution node number.

7. The optimization method for cost reduction and efficiency improvement based on pilot according to claim 6, characterized in that, Adaptive adjustment of the process parameters for the pilot project, including: The optimization scheme file is converted into a structured instruction set, in key-value pair format; The instruction set is uploaded to the configuration receiving module of the management and control system via the application interface; The configuration receiving module parses the contents of the instruction set and generates a control task sequence; The execution module modifies the process parameters one by one to the corresponding target values ​​according to the control task sequence; Then record the modified parameter status to the runtime log file.

8. The optimization method for cost reduction and efficiency improvement based on pilot according to claim 7, characterized in that, The operational data after implementation will be collected within a preset period of 7 days. The operational data after implementation specifically includes: Automatically extract the operation records of each production cycle after the cycle ends; Data collected includes actual process parameters, unit product cost, total output, yield rate, average production cycle time, resource utilization rate, total energy consumption, and number of alarm events. The collected data from the seven cycles were merged in chronological order to generate a complete post-implementation operational data table; Perform standardization processing on the post-implementation running data table, consistent with the sample dataset, and output a standardized running dataset.

9. The optimization method for cost reduction and efficiency improvement based on pilot according to claim 8, characterized in that, Based on the operational data after implementation, the cost reduction and efficiency improvement indicators of the parameter group are calculated, including: Read the unit cost data before and after implementation from the standardized operation dataset, and calculate the cost reduction index C. The calculation formula is as follows: in, The unit product cost before implementation. This refers to the unit product cost after implementation. Read the average production cycle time, yield rate, and resource utilization rate data before and after implementation from the standardized operation dataset, and calculate the efficiency improvement index E. The calculation formula is as follows: in, and These represent the average production cycle time before and after implementation. and The yield rates before and after implementation are respectively. and These represent resource utilization rates before and after implementation. , , These are the weighting coefficients for average production cycle time, yield rate, and resource utilization rate, respectively. The cost reduction index C and efficiency improvement index E are output as input results for threshold comparison.

10. The optimization method for cost reduction and efficiency improvement based on pilot according to claim 9, characterized in that, The method for setting the first threshold and the second threshold includes: Retrieve statistical data on cost reduction and efficiency improvement indicators from at least 10 similar pilot projects from the historical project database; Calculate the average of historical cost reduction indicators and multiply it by a correction factor of 1.05 to obtain the first threshold. Calculate the upper quartile of the historical efficiency improvement index and multiply it by a correction factor of 1 to obtain the second threshold; The first threshold and the second threshold are recorded in the threshold configuration table as fixed inputs to the threshold comparison module.

Images