A multi-station multi-project test scheduling method for gas meter measurement verification line

Abstract

The application discloses a kind of gas meter measurement verification line multi-station multi-project test scheduling method, comprising: according to the resource condition of gas meter measurement verification line verification, the scheduling demand in the process of gas meter verification is analyzed;Build gas meter measurement verification flexible job shop scheduling problem mathematical model, the overall processing scheduling strategy is formed by multiple gas meter scheduling information;Build gas meter measurement verification scheduling information particle coding;Using discrete particle swarm optimization algorithm, the scheduling problem is optimized and solved, the algorithm iteration parameter is set, and particle swarm optimization algorithm iteration solution is carried out;Output gas meter measurement verification scheduling information and processing completion time, and draw gas meter measurement verification scheduling Gantt chart.

Description

Technical Field

[0001] This invention relates to the fields of metrological verification and intelligent scheduling technology, and in particular to a multi-station, multi-item testing and scheduling method for a gas meter metrological verification line. Background Technology

[0002] A gas meter is an automatic mechanical instrument that uses a flexible diaphragm metering chamber to measure gas volumetric flow rate. The National Metrological Verification Regulation JJG577-2012 stipulates that gas meters using natural gas or liquefied petroleum gas as the medium generally have a service life of no more than 10 years and require periodic mandatory replacement, resulting in a large workload for verification. Furthermore, the verification regulations for gas meters specify numerous items and have complex requirements. Therefore, how to intelligently schedule the gas meter verification process and improve its efficiency and automation level has become an urgent problem to be solved.

[0003] A multi-station, multi-item testing scheduling method for gas meter calibration lines mainly includes calibration line optimization methods and scheduling solution methods based on intelligent optimization algorithms. Calibration line optimization methods, such as patents CN114705274A and CN108957385A, employ flexible fixing devices to secure the products under test. Scheduling optimization solution methods, such as CN113377073A and CN113807022A, utilize multi-agent systems and BP neural networks for scheduling optimization.

[0004] The specific patent prior art documents mentioned above are as follows:

[0005] 1) "A Fixing Device and Method for Diaphragm Gas Meters Applicable to Automatic Verification Lines," Patent No. CN114705274A. This invention discloses a fixing device and method for diaphragm gas meters applicable to automatic verification lines. The device includes: a diaphragm gas meter adapter for adapting and fixing the diaphragm gas meter; a tray on which the diaphragm gas meter adapter is mounted; a detection component for detecting whether the diaphragm gas meter and the diaphragm gas meter adapter are compatible; and a steering component for rotating the position of the tray to ensure that the diaphragm gas meter adapter and the diaphragm gas meter are compatible and fixed. The fixing device for diaphragm gas meters provided by this invention can flexibly adjust the orientation of the fixing device according to different gas intake methods of the diaphragm gas meter to adapt to the structure of the diaphragm gas meter, thereby achieving the fixing of the diaphragm gas meter and completing the automatic verification of the diaphragm gas meter. The above invention uses a mechanical device to complete the fixing and verification optimization of the gas meter. The scheduling verification optimization method of this invention is different from the above invention.

[0006] 2) "A Method and Apparatus for Confirming Abnormal Meter Positions in an Automated Verification Line for Electricity Metering Equipment", Patent No. CN108957385A. This invention provides a method and apparatus for confirming abnormal meter positions in an automated verification line for electricity metering equipment. The method includes: S1: Obtaining a sliding window of verification data for the current meter position at the current time point, where the verification data is the verification result of the current meter position on the electricity metering equipment; S2: Calculating an alarm coefficient based on the verification data corresponding to each time point within the sliding window of the verification data, determining whether the alarm coefficient is greater than a preset first threshold. If yes, proceed to S3; otherwise, proceed to S4; S3: Determining the current meter position as a suspected abnormal meter position, calculating the trend of the verification pass rate of the current meter position within a preset time period, determining whether the trend of the verification pass rate is greater than a preset second threshold. If yes, determining the suspected abnormal meter position as an abnormal meter position; otherwise, proceed to S4; S4: Determining the next meter position as the current meter position, re-executing S1, until all meter positions are confirmed at the current time point. This invention differs from the above in that it constructs a mathematical model for the scheduling of the calibration line and solves it using optimization methods, thereby optimizing the calibration efficiency of the automated calibration line, which is significantly different from the inventions described above.

[0007] 3) "A Flexible Job Shop Scheduling Optimization Method Based on a Two-Layer Multi-Agent System", Patent No. CN113377073A. This invention discloses a flexible job shop scheduling optimization method based on a two-layer multi-agent system. Specifically, it introduces the concept of service units and considers machine failures and random workpiece arrivals, establishing a dynamic scheduling model for a flexible job shop with minimum completion time. Secondly, it designs a two-layer multi-agent system architecture solution model, including a rule-based top-level scheduling mechanism and a negotiation-based bottom-level scheduling mechanism. The upper layer of the multi-agent system is the top-level scheduling of the system, where the workshop agent allocates tasks to each service unit through communication and negotiation between the workshop agent, unit agent, and workpiece agent. The lower layer of the multi-agent system is the bottom-level scheduling of the system, where unit agents allocate tasks to each machine. This invention improves solution efficiency, enables rapid response to interference problems, greatly enhances the robustness of the scheduling system, improves production efficiency, and brings greater benefits to enterprises. This invention differs from the inventions described above. In this invention, the gas meter calibration line has the characteristics of one-time arrival of workpieces, no consideration of machine failures, and the processing route can be non-fixed. The flexible workshop scheduling mathematical model constructed based on these characteristics is different, and intelligent optimization algorithms are used for scheduling optimization, which is significantly different from the inventions described above.

[0008] 4) "An Active Power Scheduling Solution Based on Pigeon Flock Algorithm and BP Cost Prediction Network", Patent No. CN113807022A. This invention belongs to the field of power system optimization scheduling technology, involving power flow calculation, intelligent algorithm improvement, and neural network applications. Specifically, it relates to an active power scheduling solution based on pigeon flock algorithm and BP cost prediction network. The method improves the performance of the basic pigeon flock algorithm in solving the multi-objective active power scheduling problem of the power grid by using a disadvantaged solution update strategy and nonlinear weight coefficients, successfully obtaining a uniformly distributed Pareto non-dominated solution set and finding the optimal compromise scheduling scheme. This invention also constructs a fuel cost prediction network based on a BP neural network. By inputting the compromise scheduling scheme into the BP neural network, multiple high-quality multi-objective active power scheduling schemes for the power grid can be determined with relatively low time cost, effectively reducing fuel costs, power losses, and exhaust emissions in power grid operation, thereby meeting the diverse needs of decision-makers. This invention differs from the above inventions, which use a scheduling solution method combining intelligent optimization algorithms and BP neural networks. This invention uses a discrete particle swarm optimization algorithm to optimize the scheduling problem, resulting in better convergence and solution efficiency. Summary of the Invention

[0009] To address the aforementioned technical problems, the purpose of this invention is to provide a multi-station, multi-item testing scheduling method for gas meter calibration lines.

[0010] The objective of this invention is achieved through the following technical solution:

[0011] A method for scheduling multi-station, multi-item testing on a gas meter calibration line, comprising:

[0012] A. Analyze the scheduling needs during the gas meter calibration process based on the calibration resource availability of the gas meter calibration line;

[0013] B. Construct a mathematical model for the scheduling problem of flexible operation workshop for gas meter metering verification, which consists of the overall processing scheduling strategy composed of scheduling information from multiple gas meters;

[0014] C. Constructing particle coding for gas meter metering verification and scheduling information;

[0015] D. Use the discrete particle swarm optimization algorithm to optimize and solve the scheduling problem, set the algorithm iteration parameters, and perform iterative solution using the particle swarm optimization algorithm;

[0016] E. Output gas meter calibration and verification scheduling information and processing completion time, and draw a Gantt chart of gas meter calibration and verification scheduling.

[0017] Compared with the prior art, one or more embodiments of the present invention may have the following advantages:

[0018] By analyzing the structure of gas meter calibration lines, the existing scheduling requirements and optimization objectives can be derived. A calibration scheduling model based on the flexible job shop scheduling problem is developed, demonstrating reliability and robustness. The discrete particle swarm optimization algorithm is used to optimize the solution of the scheduling problem, exhibiting good convergence and high solution efficiency. This method has significant academic value and practical implications for improving the testing efficiency and automation level of gas meter calibration lines. Attached Figure Description

[0019] Figure 1 This is a flowchart of the multi-station, multi-item testing and scheduling method for gas meter calibration lines;

[0020] Figure 2 This is a flowchart of the gas meter metering verification process, which outlines the multi-station, multi-item testing and scheduling method for gas meter metering verification lines.

[0021] Figure 3 It is a discrete particle coding diagram of the multi-station, multi-item test scheduling method for gas meter calibration lines;

[0022] Figure 4 It is a discrete particle coding graph scheduling Gantt chart for multi-station, multi-item testing scheduling methods in gas meter metering and calibration lines. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be described in further detail below with reference to the embodiments and accompanying drawings.

[0024] like Figure 1 The diagram shown is a flowchart of the multi-station, multi-item test scheduling method for a gas meter calibration line.

[0025] Step 10: Analyze the scheduling needs during the gas meter calibration process based on the calibration resource availability of the gas meter calibration line.

[0026] Step 20: Mathematical model of the flexible workshop scheduling problem for gas meter metering verification. The scheduling information for the i-th gas meter is denoted as T. i Including the j-th process P i,j The corresponding processing start time S i,j Processing completion time E i,j and P i,j The assigned machine r i,j The overall processing and scheduling strategy T is composed of scheduling information from multiple gas meters. P ;

[0027] Step 30 constructs a particle code for gas meter metering verification scheduling information based on a discrete coding method of machine selection and process arrangement. Each particle code consists of a process sequence code OS and a machine selection code MA, which can be decoded into an overall processing scheduling strategy T.P ;

[0028] Step 40 uses the Discrete Particle Swarm Optimization (DPSO) algorithm to optimize and solve the scheduling problem, sets the algorithm iteration parameters, and uses the processing scheduling strategy T. P The reciprocal of the maximum completion time is used as the fitness function for iterative algorithm solution;

[0029] Step 50: The discrete particle swarm optimization algorithm is completed, and the gas meter metering verification and scheduling information T is output. i and processing completion time E i,j Draw a Gantt chart for the gas meter metering verification and scheduling.

[0030] Step 10 above specifically includes: such as Figure 2 As shown, the gas meter calibration line consists of an appearance inspection device, an additional function inspection device, a sealing inspection device, and a nozzle-type gas meter calibration device. The gas meter calibration line has low testing efficiency due to long waiting times for some calibration items. Based on the scheduling needs of the calibration line, the calibration process is intelligently scheduled.

[0031] Step 20 above specifically includes: scheduling gas meter metering verification based on the flexible workshop scheduling problem, where the scheduling information for the i-th gas meter is set as T. i Including the j-th process P i,j The corresponding processing start time S i,j Processing completion time E i,j and P i,j The assigned machine r i,j The overall processing and scheduling strategy T is composed of scheduling information from multiple gas meters. P ;

[0032]

[0033] T P ={T1,T2,...T i ...,T l} (2)

[0034] A gas meter automatic calibration line currently has a batch of gas meters awaiting inspection, consisting of 2 boxes, with 6 gas meters in each box, totaling 12 gas meters. The gas meter calibration resources are shown in Table 1. These gas meters will undergo tests on appearance, sealing, pressure loss, indication error, and the function of additional devices. The number of machines available on the calibration line and the corresponding testing time vary depending on the specific calibration item.

[0035] Table 1. Gas Meter Calibration Resources

[0036]

[0037] The scheduling problem is modeled based on flexible job shop scheduling, and the optimization objective function of this problem is the maximum makespan of the scheduling strategy T P :

[0038]

[0039] According to the existing process resource situation, the following model constraints are established: ① The detection start time is less than the completion time; ② The processes are carried out according to a fixed route. If j1<j2, it means that process 1 has a higher priority than process 2, then the completion time of process 1 is less than the start time of process 2 ③ At the same moment, only one process test is performed on one verification device; ④ For the same gas meter, only one process test is performed at the same moment. Constraints ①, ②, ③, and ④ can be expressed by equations (4), (5), (6), and (7) respectively:

[0040]

[0041]

[0042]

[0043]

[0044] The above step 30 specifically includes: using a discrete coding method based on machine selection and process arrangement, each particle code consists of an Operation Sequence (OS) code and a Machine Assignment (MA) code, both of which have a length equal to the sum of the number of processes of all gas meters, and can be decoded into the overall processing scheduling strategy T P . Assume that the code of an instance is OS={1, 2, 1, 1, 2}, MA={3, 1, 2, 2, 3}. In OS, "1" and "2" represent gas meters J1 and J2 respectively, and the occurrence count j indicates the j-th verification process P of gas meter i i,j , and the verification start time S of the gas meter is arranged according to the order of OS from left to right i,j ; From left to right, MA respectively represents the machine selected for process P i,j , the first code "3" represents that process P 1,1 is processed on the third verification device (M4) in the optional machine set (as Figure 3 shown).

[0045] The above step 40 specifically includes: using the Discrete Particle Swarm Optimization (DPSO) algorithm to solve the scheduling problem, and using the discrete particle update formula (8) to update the particles:

[0046]

[0047] In the formula, c1 is the individual optimization learning factor, c2 is the population optimization learning factor, f1 and f2 are the particle operation operators, f1 is the crossover operation between two particles, f2 is the mutation operation of a single particle, and w is the inertia weight, representing the mutation probability of a single particle. This represents an operational relationship, where 'a' is a uniformly random number within the range [0,1].

[0048]

[0049]

[0050] Set the algorithm iteration parameters: maximum number of iterations G max Given the particle swarm size N, individual optimization learning factor c1, overall optimization learning factor c2, and inertia weight w, this algorithm optimizes the above scheduling problem. Algorithm 1 is the discrete particle swarm optimization algorithm flow. The algorithm first initializes the population and sets the initial position x of each particle. i The encoding process is performed, and the fitness F(i) is calculated. In each subsequent iteration, the position of each particle is updated using Equation (8), and the fitness function is evaluated to update the individual's historical best position. and the global optimal position After the iteration, the algorithm decodes and outputs the optimal overall processing scheduling strategy T. P .

[0051]

[0052]

[0053] Step 50 above specifically includes: after the discrete particle swarm optimization algorithm is completed, outputting the gas meter metering verification and scheduling information T. i and processing completion time E i,j According to the gas meter metering verification and dispatch information T i Draw a Gantt chart for gas meter calibration and scheduling, such as... Figure 4 As shown, the scheduling Gantt chart includes the machines (r) arranged for each process of the gas meter. k The j-th process P i,j The corresponding processing start time S i,j Processing completion time E i,j Scheduling strategy T P Maximum completion time min(T) P ).

[0054] While the embodiments disclosed in this invention are as described above, the content is merely for the purpose of facilitating understanding of the invention and is not intended to limit the invention. Any person skilled in the art to which this invention pertains may make any modifications and variations in form and detail of the implementation without departing from the spirit and scope disclosed herein; however, the scope of patent protection for this invention shall still be determined by the scope defined in the appended claims.

Claims

1. A method for scheduling multi-station, multi-item testing on a gas meter calibration line, characterized in that, include: A. Analyze the scheduling needs during the gas meter calibration process based on the calibration resource availability of the gas meter calibration line; B. Construct a mathematical model for the scheduling problem of flexible operation workshop for gas meter metering verification, which consists of the overall processing scheduling strategy composed of scheduling information from multiple gas meters; C. Constructing particle coding for gas meter metering verification and scheduling information; D. Use the discrete particle swarm optimization algorithm to optimize the scheduling problem, set the algorithm iteration parameters, and use a processing scheduling strategy. The reciprocal of the maximum completion time is used as the fitness function for iterative solution using the particle swarm optimization algorithm; E. Output gas meter metering verification and scheduling information and processing completion time, and draw a Gantt chart of gas meter metering verification and scheduling; In step A, the gas meter calibration line consists of an appearance inspection device, an additional function inspection device, a sealing inspection device, and a nozzle-type gas meter calibration device; the gas meter is tested for appearance, sealing, pressure loss, indication error, and additional device functions. In section B, gas meter calibration scheduling is based on the flexible workshop scheduling problem. Individual gas meter dispatch information set , including the Item process Corresponding processing start time Processing completion time as well as Assigned machines Multiple gas meter scheduling information forms the overall processing and scheduling strategy. : （1）； （2）； The gas meter calibration and scheduling problem is modeled based on flexible job shop scheduling. The optimization objective function of the gas meter calibration and scheduling problem is to minimize the scheduling strategy. Maximum completion time: （3）； Based on the existing process resources, the following model constraints are established: (1) If the detection start time is less than the completion time, the expression is: （4）； (2) The process is carried out along a fixed route, if This indicates that process 1 has a higher priority than process 2, therefore the completion time of process 1 is... Less than the start time of process 2 : （5）； (3) When only one process test is performed on a single calibration device at the same time: （6）； (4) When only one process is tested for the same gas meter at the same time: （7）； In C, a discrete coding method based on machine selection and process arrangement is used. Each particle code consists of a process sorting code OS and a machine selection code, both of which have a length equal to the sum of the number of processes in all gas meters. This is decoded into an overall processing scheduling strategy. ; In the D section, the discrete particle swarm optimization algorithm is used for scheduling and the discrete particle update formula (8) is used to update the particles: （8） In the formula, To optimize the learning factor for individuals, To optimize the learning factor for the population, , For particle operation operators, This is a crossover operation between two particles. For mutation operations on a single particle, The inertial weights represent the mutation probability of a single particle. This represents an operational relationship. a Uniformly random numbers in the range [0,1]: （9）； （10）； Set algorithm iteration parameters: maximum number of iterations Particle swarm size Individual Optimization Learning Factor Overall optimization of learning factors Inertia weights To optimize the solution of the scheduling problem, the algorithm first initializes the population and sets the initial positions of each particle. Encoding, calculating fitness The fitness function is used to evaluate the quality of the scheduling scheme corresponding to the particles. In each subsequent iteration, the position of each particle is updated by Equation (8), and the fitness function is evaluated to update the individual's historical best position. and the global optimal position After the iteration, the algorithm decodes and outputs the optimal overall processing scheduling strategy. .

2. The multi-station, multi-item test scheduling method for gas meter calibration lines as described in claim 1, characterized in that, In step A, the verification process is intelligently scheduled according to the scheduling needs of the verification line.

3. The multi-station, multi-item test scheduling method for gas meter calibration lines as described in claim 1, characterized in that, In E, after the discrete particle swarm optimization algorithm is completed, it outputs gas meter metering verification and scheduling information. and processing completion time According to the gas meter metering verification and dispatch information Draw a Gantt chart for the gas meter calibration and verification process. The Gantt chart includes the machines arranged for each process of the gas meter calibration. , No. Item process Corresponding processing start time Processing completion time Minimize scheduling strategy Maximum completion time .

Images