Equipment rotation decision-making method, equipment, medium and product

Abstract

The invention discloses an equipment rotation decision-making method, equipment, a medium and a product, and relates to the field of data processing. The method comprises the following steps: replacing a fuzzy number of a rule antecedent in a belief rule base with a language Z number to construct a rotation decision model; obtaining a historical operation task and an optimal rotation strategy corresponding to the historical operation task; taking the historical operation tasks and the optimal rotation strategy corresponding to the historical operation tasks as training data to optimize the rotation decision model; acquiring a current operation task of the equipment; and for the current operation task, reasoning is carried out by adopting the optimized rotation decision model, and an optimal rotation strategy is obtained. According to the method, the reasonability and accuracy of decision making can be improved, and decision making technical support is provided for long-time and large-scale operation of equipment.

Description

Technical Field

[0001] The present application relates to the field of data processing, and in particular to a device rotation decision method, device, medium and product. Background Art

[0002] Long-term, large-scale operation of industrial equipment has become a common working mode. For example, a group of oil and gas storage tanks need to work collectively over a period of time to complete the loading and unloading tasks of oil and gas. During the operation of the equipment, due to long-term power supply, natural wear and tear, and environmental interference, its health status will decline, resulting in an increased probability of failure. Therefore, after a period of operation, the equipment needs to be withdrawn from operation for maintenance, and other idle equipment is used to replace it. In this way, the problem of equipment rotation operation arises.

[0003] In order to formulate equipment rotation plans, three equipment rotation strategies have been formed in the project, namely, single-line single-equipment strategy, single-line multi-equipment strategy, and multi-line multi-equipment strategy. These rotation strategies have their own advantages. Among them, the single-line single-equipment strategy uses fewer equipment and has low costs, but its rotation plan has a high risk. The single-line multi-equipment strategy has a large number of equipment and high costs, but it brings a relatively low risk. At present, there is no scientific and objective method or mechanism to select and make decisions on rotation strategies in engineering, and the selection of rotation strategies is arbitrary and subjective. Therefore, for an equipment operation task, how to reasonably select the best rotation strategy is an urgent problem to be solved in rotation decision-making.

[0004] The selection of the above equipment rotation plan is essentially a multiple attribute decision making (MADM) problem, that is, to sort and select a limited number of alternative plans based on multiple attributes and make decisions. In engineering, the quality of a rotation plan is generally judged by three attributes: economic cost, administrative cost and risk. The economic cost comes from the cost of equipment maintenance, so it directly depends on the number of equipment used. The administrative cost comes from the equipment management registration procedures in the project. When the equipment is replaced from one operating line and transferred to another operating line, re-approval, registration and other operations are required, which need to be transmitted on the administrative link. Too many switches between operating lines will bring high administrative costs. The risk level is comprehensively reflected by the health status of the operating equipment.

[0005] Over the past decades, many MADM methods have been developed, such as Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS), grey correlation analysis, TOPIM (TOmada de Decisão Iterativa Multicritério) method, and analytic hierarchy process. These methods have the following two problems when applied to equipment rotation decisions: First, these methods are affected by the arbitrariness of experts' subjective knowledge, and have strong subjectivity in weight setting and decision-making process, resulting in inaccurate results. Second, in the rotation decision, the risk of the current rotation plan is reflected by the health status of the equipment in the future. For long-term operations, it is difficult to accurately predict the health change trend of the equipment, which makes it difficult to directly obtain the risk of the rotation plan.

[0006] Belief rule base (BRB) is a decision model based on IF-THEN rules. Its outstanding feature is that the rule consequent is described as the confidence distribution of all possible results. Therefore, BRB can describe subtle changes in human expert knowledge. For decades, BRB has been widely used in the field of decision-making. In the equipment rotation decision problem, the rule antecedent of BRB is the specific attributes of the task, such as task duration and different operation routes, and the rule consequent is three alternative rotation strategies.

[0007] Although BRB does not need to quantitatively calculate the risk of the rotation scheme and can effectively use the knowledge and experience of human experts in equipment rotation decisions to construct a decision model, it cannot represent the impact of unreliable knowledge on the decision. Due to the subjectivity and uncertainty of human knowledge, expert knowledge is inherently unreliable. Specifically, experts have different proficiency in different knowledge, and the decision-making process driven by unreliable knowledge may bring greater risks and cause loss of life and property. Therefore, it has become a technical problem to be solved in this field to reasonably model unreliable knowledge in the BRB method and quantitatively analyze its impact on decision results. Summary of the invention

[0008] The purpose of this application is to provide an equipment rotation decision-making method, equipment, medium and product, which can improve the rationality and accuracy of decision-making and provide decision-making technical support for long-term and large-scale operation of equipment.

[0009] To achieve the above objectives, this application provides the following solutions: In a first aspect, the present application provides a device rotation decision method, comprising: The fuzzy numbers of rule antecedents in the confidence rule base are replaced with linguistic Z numbers to construct a rotation decision model; Obtaining historical job tasks and optimal rotation strategies corresponding to the historical job tasks; Using the historical operation tasks and the optimal rotation strategy corresponding to the historical operation tasks as training data to optimize the rotation decision model; Get the current job task of the device; For the current task, the optimized rotation decision model is used for reasoning to obtain the optimal rotation strategy.

[0010] Optionally, the rule consequence in the rotation decision model is a rotation strategy to be selected; the rotation strategy to be selected includes one or more of a single-line single-device strategy, a single-line multi-device strategy and a multi-line multi-device strategy.

[0011] Optionally, the constructed rotation decision model is expressed as:

[0012] In the formula, is the operation duration of the task, is the number of operation lines of the operation task, The duration of the task The corresponding language Z number, The number of operation lines for the operation task The corresponding language Z number, is the confidence of the single-line single-device strategy, is the confidence of the single-line multi-device strategy, is the confidence of the multi-line multi-device strategy, is the rule reliability, For a single-line single-device strategy, For a single line multi-device strategy, For multi-line multi-device strategy, is the kth language Z-number rule, Indicates the assumption, Indicates yes, Indicates connection, THEN indicates the result, Indicates the reliability of a rule.

[0013] Optionally, the rule reliability is jointly determined by the reliability constraints of all language Z numbers in a rule; wherein the formula for determining the rule reliability is expressed as: ; In the formula, is the real value corresponding to the reliability constraint in the kth rule of the ith indicator, is the lower bound of the rule reliability, is the total number of indicators.

[0014] Optionally, for the current task, the optimized rotation decision model is used for reasoning to obtain the optimal rotation strategy, which specifically includes: Obtaining the operation duration and the number of operation lines of the current operation task; Determining the activation degree of each rule based on the operation duration and the number of operation lines of the current operation task; activating the language Z-number confidence rules in the optimized rotation decision model based on the activation degree of each rule; when the language Z-number confidence rules are activated, each language Z-number confidence rule includes rule reliability, rule activation weight and rule consequent confidence distribution; Designing an extended proposition evidence reasoning rule, and introducing an extended proposition into the extended proposition evidence reasoning rule; the extended proposition is used to accommodate the unreliability of evidence; Based on the extended proposition evidential reasoning rule with the extended proposition introduced, the discounted confidence is obtained by discounting the confidence distribution of the language Z-number confidence rule after the rule reliability is activated; Adopting an analytical evidence reasoning algorithm to fuse the discounted confidence and obtain a fusion result; Performing anti-discounting processing on the fusion result to obtain the confidence distribution and result reliability of the rotation strategy to be selected; The optimal rotation strategy is obtained based on the confidence distribution of the rotation strategy to be selected and the reliability of the result.

[0015] Optionally, using the historical operation tasks and the optimal rotation strategy corresponding to the historical operation tasks as training data to optimize the rotation decision model specifically includes: Constructing a parameter optimization model of the rotation decision model based on the training data; The particle swarm algorithm is used to search for the optimal solution of the parameter optimization model, and the optimal solution is used as the optimal parameter of the rotation decision model to complete the optimization of the rotation decision model.

[0016] Optionally, the parameter optimization model of the rotation decision model is expressed as:

[0017] In the formula, is the cross entropy objective function, is the discriminant variable, min means taking the minimum value, is the confidence of the nth rotation strategy to be selected, is the operation duration of the task, is the number of rotation strategies to be selected, is the number of operation lines of the operation task, is the number of data sets, is the confidence of the n1th result in the sth data, is the label value of the sth data, Indicates conditions, Represents the discriminant variable of the i-th indicator.

[0018] In a second aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the device rotation decision method provided above.

[0019] In a third aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the device rotation decision method provided above.

[0020] In a fourth aspect, the present application provides a computer program product, including a computer program, which implements the device rotation decision method provided above when executed by a processor.

[0021] According to the specific embodiments provided in this application, this application discloses the following technical effects: The present application provides a method, device, medium and product for equipment rotation decision-making. By replacing the fuzzy number of the rule antecedent in the confidence rule base with the language Z number to construct a rotation decision model, the problem of incomplete reliability of expert knowledge can be solved. Using historical work tasks and the optimal rotation strategy corresponding to the historical work tasks as training data to optimize the rotation decision model can avoid the problem of inaccurate and unreasonable decision results caused by the subjectivity and uncertainty of expert knowledge, thereby improving the rationality and accuracy of the decision. For the current work task, the optimized rotation decision model is used for reasoning to obtain the optimal rotation strategy, which can provide decision-making technical support for long-term and large-scale operations of equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0023] Figure 1 This is an application environment diagram of a device rotation decision method in an embodiment of the present application; Figure 2 A schematic diagram of a process flow of a device rotation decision method provided in one embodiment of the present application; Figure 3 A schematic diagram of a rotation operation of 50 historical operation tasks provided in another embodiment of the present application; Figure 4 A schematic diagram of an optimal rotation strategy for 50 historical job tasks provided in another embodiment of the present application; Figure 5 A schematic diagram of a decision result of LZ-BRB after training provided in another embodiment of the present application; Figure 6 A schematic diagram of the reliability of a decision result provided by another embodiment of the present application; Figure 7 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0025] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0026] The device rotation decision method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the current job task of the device to the server 104. After the server 104 receives the current job task of the device, for the current job task of the device, the server 104 uses the optimized rotation decision model to reason and obtain the optimal rotation strategy. Among them, the fuzzy number of the rule antecedent in the confidence rule base is replaced with the language Z number to construct a rotation decision model. The historical job tasks and the optimal rotation strategy corresponding to the historical job tasks are used as training data to optimize the rotation decision model. The server 104 can feedback the obtained optimal rotation strategy to the terminal 102. In addition, in some embodiments, the device rotation decision method can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly determine the optimal rotation strategy for the current job task of the device, or the server 104 can obtain the current job task of the device from the data storage system and determine the optimal rotation strategy for the current job task of the device.

[0027] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or may be a cloud server.

[0028] In an exemplary embodiment, Figure 2 As shown, a device rotation decision method is provided, which is executed by a computer device, and can be specifically executed by a computer device such as a terminal or a server alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used as an example to illustrate, including the following steps 200 to 204. Among them: Step 200: Replace the fuzzy number of the rule antecedent in the confidence rule base with the linguistic Z number to construct a rotation decision model.

[0029] Step 201: Obtain historical job tasks and the optimal rotation strategy corresponding to the historical job tasks.

[0030] Step 202: Use the historical operation tasks and the optimal rotation strategies corresponding to the historical operation tasks as training data to optimize the rotation decision model.

[0031] Step 203: Get the current job task of the device.

[0032] Step 204: For the current task, the optimized rotation decision model is used for reasoning to obtain the optimal rotation strategy.

[0033] The implementation of the above steps 200 to 204 can improve the rationality and accuracy of decision-making and provide technical support for decision-making for long-term and large-scale operation of equipment.

[0034] In another exemplary embodiment of the present application, in order to reasonably model unreliable knowledge in the BRB method and quantitatively analyze the impact of unreliable knowledge on decision-making results. In this embodiment, in the long-term rotation of equipment, experts have formed empirical knowledge about the relationship between work tasks and the rotation strategies to be selected. Generally speaking, work tasks can be reflected by two attributes: work duration and the number of work lines. The work duration affects the equipment rotation in the time dimension. The longer the work duration, the longer the rotation duration and number of equipment, and the economic cost, administrative cost and risk of the entire work task will increase. The work line affects the equipment rotation from the spatial dimension. The number of work lines determines the intensity of the equipment rotation. Therefore, in the equipment rotation decision problem, a work task is regarded as the problem input, and the specific input variables are the work duration and the number of work lines.

[0035] Due to the subjectivity and uncertainty of human knowledge, expert knowledge is not completely reliable, which is reflected in their different levels of proficiency in different types of knowledge.

[0036] Specifically, experts have different proficiency in different value ranges of variables. Experts will be more familiar with values ​​that appear frequently in indicators. Based on this, the linguistic Z-number is used to model unreliable expert knowledge. The linguistic Z-number (LZ) is a binary array expressed as: (1) In the formula, Indicates the language Z number, is the fuzzy constraint of the linguistic variables, is the reliability constraint of the first component. For example, if the expert is less familiar with the task duration when it is shorter and more familiar with it when it is longer, the language Z-number can be expressed as: ("shorter", "less familiar") and ("longer", "more familiar").

[0037] Based on the above-given linguistic Z-number, in order to reasonably model the empirical knowledge of human experts about the task and the rotation strategy to be selected, and reflect its unreliability, in this embodiment, the rotation decision model constructed in the above step 200 of the present application is a linguistic Z-number belief rulebase (LZ-BRB) model. Among them, the fuzzy number of the rule antecedent in the original BRB is replaced by the linguistic Z-number to simultaneously represent the fuzziness and reliability of the expert.

[0038] In the LZ-BRB model, the rule consequences are three rotation strategies to be selected, namely, the single-line single-device strategy, the single-line multi-device strategy, and the multi-line multi-device strategy. The LZ-BRB model can be expressed as: (2) in, They are respectively the operation duration and the number of operation lines of the operation task. are the language Z numbers corresponding to the two input variables respectively. , and The confidence levels of the single-line single-device strategy, the single-line multi-device strategy, and the multi-line multi-device strategy reflect the support level of the rules for each of them. Rule reliability is generated from unreliable knowledge and reflects the ability of a rule to provide reliable information. is the kth language Z-number rule, Indicates the assumption, Indicates yes, Indicates connection, THEN indicates the result, Indicates the reliability of a rule. For a single-line single-device strategy, For a single line multi-device strategy, For multi-line multi-device strategies, the specific implementation process of these three strategies is as follows: (1) Single-line single-device strategy : exist In the process, one device always works for a long time on one operation line. When a device needs to be maintained due to health degradation, other idle devices will temporarily replace it. After the maintenance is completed, the device will continue to work. The characteristics are that it uses less equipment, has a simple implementation process, does not involve switching of equipment on multiple lines, and has a high risk.

[0039] (2) Single-line multiple-device strategy : exist In a work line, two or more devices share the working time. When a device is out of work due to health decline, another device will take over for a long time, and the process will be repeated in this way. The characteristics are that it uses more equipment, the implementation process is relatively simple, and the risk is relatively low.

[0040] (3) Multi-line and multi-device strategy : exist In the process, the equipment is not limited to a certain operation line. When an operation line is withdrawn and maintenance is completed, the equipment will be used for the operation line that needs to be replaced. The characteristics are that less equipment is used and the risk level is lower than , higher than . However, the implementation process is relatively complicated and involves the switching of equipment between different operating lines.

[0041] The rule reliability in the LZ-BRB model is determined by the reliability constraints of all the linguistic Z numbers in a rule, which can be calculated as follows: (3) (4) in, is the real value corresponding to the reliability constraint in the kth rule of the ith indicator, is the lower bound of the rule reliability, is the total number of indicators. is the reliability measure of the i-th input variable in the k-th rule. means represents a mapping.

[0042] Based on the above description, the LZ-BRB model can reasonably model the empirical knowledge of human experts about the task and the rotation strategy to be selected, and reflect its unreliability.

[0043] For example, for a long-duration equipment operation task with many operation lines, experts prefer to use and Strategy. The rules can be expressed as: (5) In another exemplary embodiment of the present application, in order to realize decision result reasoning based on extended propositional evidential reasoning rules, the above-mentioned LZ-BRB model represents human experience knowledge in equipment rotation decision-making. For a newly acquired work task (i.e., the current work task of the equipment), the constructed LZ-BRB model can be used to reason and obtain its optimal rotation strategy.

[0044] In order to quantitatively reflect the impact of unreliable knowledge on decision-making results, this embodiment designs an extended propositional evidence reasoning rule (ER-EP) rule for the reasoning of rules in the LZ-BRB model. The newly obtained task is recorded as ,in is the duration of the task, is the number of operating lines. The reasoning process of the optimal rotation strategy can include the following two steps: Step 1: Activation of the language Z-number confidence rule. This step requires calculating the activation degree of each rule, which depends on the matching degree between the rule antecedent and the input information. Based on this, the activation degree of each rule can be determined based on the operation duration of the current operation task and the number of operation lines. The activation weight of the rule is used to represent the activation degree. for: (6) in, represents the membership calculated by the triangular membership function, Indicates the operation duration or number of operation lines of the task.

[0045] Step 2: Reasoning of rules and obtaining decision results. Activate the language Z-number confidence rules in the optimized rotation decision model based on the activation degree of each rule. When the language Z-number confidence rules are activated, each language Z-number confidence rule contains three types of information: rule reliability, rule activation weight, and rule consequent confidence distribution. In order to fully integrate these three types of information and quantitatively characterize the impact of unreliable knowledge, ER-EP introduces an extended proposition To accommodate the unreliability of evidence, and to discount the confidence distribution through the reliability of rules, so that more reliable evidence can play a greater role in fusion. Step 2.1: Discount of basic confidence. This step is mainly based on the introduction of the extended propositional evidence reasoning rule with extended propositions. The confidence distribution of the language Z-number confidence rule activated by the rule reliability discount is used to obtain the discounted confidence. The discounted confidence distribution is calculated by the following formula: (7) (8) in, is the discounted confidence.

[0046] Step 2.2: Fusion of discounted belief distributions. The analytical evidence inference algorithm is used to fuse discounted belief distributions as follows: (9) (10) (11) (12) (13) in, , and They are all intermediate variables in the calculation. is the number of rules. =3 is the number of results. Assess confidence for intermediate processes, is the jth result.

[0047] Step 2.3: Conversion of fusion results. The fusion results in formula (9) need to be de-discounted to obtain the confidence distribution and reliability of the decision results at the same time. The confidence distribution of the decision results is expressed as: (14) (15) In the formula, represents the confidence distribution of the decision result, represents the confidence level of the intermediate process of expanding the proposition, , and Confidence of the anti-discounted single-line single-device strategy, single-line multiple-device strategy, and multiple-line multiple-device strategy.

[0048] Step 2.4: Obtain the optimal rotation strategy based on the confidence distribution of the rotation strategy to be selected and the reliability of the result. Generally, the rotation strategy with the largest confidence is taken as the optimal rotation strategy, which is expressed as : (16) Reliability of results It is expressed as: (17) Represents the confidence of the intermediate process evaluation after the fusion of the extended proposition.

[0049] In another exemplary embodiment of the present application, the above-constructed rotation decision model may produce inaccurate and unreasonable decision results due to the subjectivity and uncertainty of expert knowledge. In order to solve this problem, historical equipment rotation cases are used to train the rotation decision model, that is, the rotation decision model is optimized based on historical rotation cases and gradient descent algorithms. Specifically, a historical job task has used a certain rotation strategy to formulate an equipment rotation plan. Therefore, it is first necessary to collect job tasks that have used three rotation strategies in history. The three strategies for this task are evaluated to determine the optimal rotation strategy for the historical task, which is used as training data to optimize the rotation decision model. The various attributes of a rotation plan are: 1. Economic cost EC. The economic cost is equal to the number of equipment used for this task.

[0050] 2. Administrative cost AC. Administrative cost is equal to the number of times all equipment switches between different operating lines.

[0051] 3. Risk level RL. The risk level depends on the comprehensive health status of all equipment in the task.

[0052] Among the above three attributes, EC and AC can be calculated based on statistics. RL is obtained through the health status of the equipment. Therefore, it is necessary to first evaluate the health status of the equipment at a certain moment. In this application, a double level evidential reasoning (DLER) method is used to evaluate the quality of historical rotation plans. The DLER method calculates the scores of historical rotation plans and determines the optimal rotation strategy for a historical job task based on the scores. Specifically, in the DLER method, the first layer DLER-I evaluates the health status of the equipment and further calculates the risk of the rotation plan. The second layer DLER-II combines the risk, economic cost and administrative cost of the rotation plan to form a comprehensive score of the rotation plan.

[0053] The implementation process of the first-level DLER-I to evaluate the health status of the equipment is as follows: First, evaluate the health status of a device in operation. The health status evaluation index system of the device is constructed as follows: Based on these evaluation indicators at historical moments Data To analyze the health status of the equipment. The weight of evidence for each evaluation indicator is expressed as First, the data is transformed into a unified confidence distribution framework through a rule-based (utility) approach. The equipment health status level is The reference values ​​of each health assessment index at these levels are Therefore, the confidence distribution of each data for these health levels is expressed as: (18) in: (19) (20) (twenty one) In the formula, represents the confidence level of the kth reference level of the ith indicator, represents the confidence level of the k+1th reference level of the ith indicator, It represents the first index of the i-th index. The confidence level of the reference level.

[0054] The basic probability mass of each piece of evidence is expressed as: (twenty two) (twenty three) In the formula, represents the basic probability mass of the nth reference level of the ith indicator, represents the confidence level of the nth reference level of the ith indicator, represents the basic probability mass of global ignorance of the ith indicator, represents the weight of evidence for the ith indicator.

[0055] Then, the above is combined by the orthogonal sum formula The basic probability quality of an evaluation metric. is global ignorance. Assume and The basic probability mass after fusion is expressed as: (twenty four) (25) (26) In the formula, represents the basic probability mass of the nth reference level after integrating the k+1th rule, represents the normalization coefficient in the k+1th rule of the fusion, represents the basic probability mass of the nth reference level after integrating the kth rule, represents the basic probability mass of the nth reference level of the k+1th rule, represents the basic probability mass of global ignorance of the k+1th rule, represents the basic probability mass of global ignorance after integrating the kth rule, represents the basic probability mass of global ignorance after integrating the k+1th rule, represents the basic probability mass of the j2th reference level after integrating the k+1th rule, represents the basic probability mass of the j2th reference level after integrating the kth rule, Represents the basic probability mass of the k+1th reference level fused with the tth rule.

[0056] The combined confidence of the individual results is calculated as : (27) In the formula, represents the basic probability mass of the nth reference level after integrating M1 rules, Represents the basic probability mass of global ignorance after fusing M1 rules.

[0057] The above confidence gives the distributed evaluation result of the health status of the equipment, which needs to be converted into a single point value of health. Therefore, the expected utility formula is used to calculate the health status of the equipment at a certain historical moment. Health for : (28) in, is the utility value of the health level.

[0058] The above process realizes the control of a single device at any time. In long-term, large-scale operations with multiple devices, it is necessary to conduct health assessments on multiple devices over a period of time to reflect the risk level of the rotation plan. , expressed as: (29) in, Indicates Equipment The moment of which there are , represents the sum of the operation lines, Indicates f The number of operations of each device, Indicates the operation cycle.

[0059] The implementation process of the second layer DLER-II is: The purpose of the second layer DLER-II is to integrate the three attributes of economic cost EC, administrative cost AC and risk RL at the rotation strategy level to calculate the score of the rotation strategy and further obtain the optimal historical rotation strategy. Among them, RL is calculated by the first layer DLER-I. EC and AC respectively count the number of equipment used in a rotation plan and the total number of times the equipment switches for a long time in different operation lines. Then, the data of these three attributes are converted into a unified confidence distribution. The score reference level of the rotation strategy is The weights of the three attributes are The confidence distribution of the three attribute data is expressed as: (30) in: (31) (32) (33) In the formula, Represents the data of three attributes respectively. For the The attribute A reference value. For the The attribute A reference value. Indicates The confidence level of an attribute, represents the first index of the i-th index The confidence level of an attribute.

[0060] Then, the confidence distribution corresponding to the score reference level of the rotation strategy is calculated as: (34) in: .

[0061] In the formula, For the The weight of evidence for each attribute. is the normalization coefficient of the intermediate process. is the confidence calculated according to formula (31)-formula (33).

[0062] The expected utility formula is used to calculate the score of the rotation strategy : (35) In the formula, represents the utility of the pth reference level, represents the pth reference grade conversion score, Represents the confidence level corresponding to the p-th reference level.

[0063] The optimal historical rotation strategy is the one with the highest score, expressed as: (36) In the formula, represents the optimal historical rotation strategy.

[0064] Thus, a labeled historical case dataset can be obtained, which is expressed as: (37) in, They are the operation duration and the number of operation lines respectively. is the number of the optimal rotation strategy. The number of job tasks.

[0065] Based on the historical rotation case data set of formula (37), a parameter optimization model can be constructed to train the rotation decision model given above. The parameter optimization model is expressed as: (38) In the formula, is the cross entropy objective function, is the discriminant variable, min means taking the minimum value, is the confidence of the nth rotation strategy to be selected, is the operation duration of the task, is the number of rotation strategies to be selected, is the number of operation lines of the operation task, is the number of data sets, is the confidence of the n1th result in the sth data, is the label value of the sth data, Indicates conditions, Represents the discriminant variable of the i-th indicator.

[0066] The parameter optimization model in formula (38) is driven by the particle swarm optimization (PSO) algorithm to search for the optimal model parameter solution.

[0067] The particle swarm algorithm is used to search for the optimal solution of the parameter optimization model, and the optimal solution is used as the optimal parameter of the rotation decision model to complete the optimization of the rotation decision model.

[0068] In another exemplary embodiment of the present application, the implementation process of the method provided in the present application is described in order to determine the operation rotation decision of the oil storage tank.

[0069] The storage and loading and unloading of oil rely on oil storage tanks, which are a type of large industrial structures.

[0070] In the oil loading and unloading tasks, there are often multiple work lines within a period of time, that is, multiple oil storage tanks are working together.

[0071] This application uses the loading and unloading tasks of a certain oil receiving station as an example for practical verification.

[0072] In 2010-2024, 50 historical tasks that meet the requirements of dataset generation were selected from the receiving station. These rotating tasks are shown in Figure 3 These tasks have used three rotation strategies to generate rotation solutions.

[0073] In order to scientifically formulate a rotation plan for the current operation task, the above-mentioned language Z number BRB is used to construct an oil tank rotation decision model.

[0074] According to the experts’ experience and knowledge, the duration of the work tasks and the number of work routes are divided into four levels: {“Short” (S), “Medium” (M), “Relatively long” (RL), “Very long” (VL)} and {“Low” (Low), “Medium” (M), “Relatively high” (RH), “Very high” (VH)}. The reference values ​​corresponding to these levels are shown in Table 1.

[0075] The experts’ proficiency in these levels is described by three linguistic values: {“Unfamiliarity” (U), “Ordinarily” (O), “Familiarly” (F)}, which correspond to expert reliability of 0.5, 0.7, and 0.9, respectively. The eight linguistic Z-scores are shown in Table 2.

[0076] According to the combination of the linguistic Z numbers of the two input variables, experts can construct 4×4=16 linguistic Z number confidence rules based on their empirical knowledge, which are shown in Table 3.

[0077] Table 1 Engine operating modes and clearance reference values

[0078] Table 2 Language Z-number table

[0079] Table 3 Language Z-number confidence rules table

[0080] During the operation of oil storage tanks, settlement and deformation are two indicators that can reflect their health status. These data are also collected together to evaluate their health status. The DLER method is used to evaluate the health status of the equipment and further calculate the optimal rotation strategy for historical operation tasks, which is shown in Figure 4 The resulting data set uses the parameter optimization model in formula (35) to optimize the initial LZ-BRB model and obtain the optimized LZ-BRB model. The 50 sets of historical data are inferred by the ER-RP rule and the optimized LZ-BRB model, and the decision results are shown in Figure 5 Among them, the inference results of the initial LZ-BRB model are also Figure 5 Displayed together with. Figure 5 It can be seen that due to factors such as uncertainty and subjectivity of expert knowledge, the decision results of the initial LZ-BRB model have a certain deviation from the actual historical best rotation strategy, and the accuracy rate is only 72%. After optimization training through historical rotation cases, the optimized LZ-BRB model can more accurately judge the optimal rotation strategy, with an accuracy rate of 98%. In addition, the LZ-BRB model also provides the reliability of the decision results. According to formula (17), the reliability of the decision results of 50 historical cases is shown in Figure 6 The reliability of the results quantitatively reflects the impact of unreliable expert knowledge on the rotation decision and characterizes the credibility of the results. Results with low reliability should be used with caution in engineering. Therefore, the method proposed in this application can provide a scientific and quantitative decision-making model for the rotation decision while giving the credibility of the decision results, thereby providing valuable decision support for the formulation of decision-making plans in engineering.

[0081] In summary, the present application is applicable to the decision-making field in engineering management, and aims to make long-term and large-scale equipment operation rotation decisions under unreliable expert knowledge, determine the optimal rotation strategy, and provide decision support for scientifically and reasonably formulating rotation plans, especially involving an equipment rotation decision method driven by unreliable knowledge. The method includes the following steps: Construction of a rotation decision model based on the language Z number BRB. Reasoning of decision results based on extended propositional evidence reasoning rules. Optimization of rotation decision model based on historical rotation cases and gradient descent algorithm. Oil tank operation rotation decision based on language Z number BRB. The present application innovatively proposes a language Z number BRB model to establish an equipment rotation decision model, develops an extended propositional evidence reasoning rule to adapt the reasoning of the language Z number BRB and obtains the reliability of the decision result. The optimal rotation strategy is obtained through historical operation task evaluation to obtain a training data set, and the optimal decision model parameters are determined based on particle swarm algorithm training. The validity of the present application is verified by examples. The present application can fully consider the influence of unreliable knowledge in engineering to construct a decision model, and use historical operation cases to train it. It can analyze the reliability of the results while making decisions, and has strong engineering application potential.

[0082] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 7 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store device rotation decision data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a device rotation decision method is implemented.

[0083] Those skilled in the art will understand that Figure 7The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0084] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0085] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0086] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0087] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0088] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0089] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0090] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A device rotation decision method, characterized in that: The device rotation decision method comprises: The fuzzy numbers of rule antecedents in the confidence rule base are replaced with linguistic Z numbers to construct a rotation decision model; Obtaining historical job tasks and optimal rotation strategies corresponding to the historical job tasks; Using the historical operation tasks and the optimal rotation strategy corresponding to the historical operation tasks as training data to optimize the rotation decision model; Get the current job task of the device; For the current task, the optimized rotation decision model is used for reasoning to obtain the optimal rotation strategy.

2. The device rotation decision method according to claim 1, characterized in that: The rule consequence in the rotation decision model is the rotation strategy to be selected; the rotation strategy to be selected includes one or more of a single-line single-device strategy, a single-line multi-device strategy and a multi-line multi-device strategy.

3. The device rotation decision method according to claim 2, characterized in that: The constructed rotation decision model is expressed as: In the formula, is the operation duration of the task, is the number of operation lines of the operation task, The duration of the task The corresponding language Z number, The number of operation lines for the operation task The corresponding language Z number, is the confidence of the single-line single-device strategy, is the confidence of the single-line multi-device strategy, is the confidence of the multi-line multi-device strategy, is the rule reliability, For a single-line single-device strategy, For a single line multi-device strategy, For multi-line multi-device strategy, is the kth language Z-number rule, Indicates the assumption, Indicates yes, Indicates connection, THEN indicates the result, Indicates the reliability of a rule.

4. The device rotation decision method according to claim 3, characterized in that: The rule reliability is determined by the reliability constraints of all the language Z numbers in a rule; wherein the formula for determining the rule reliability is expressed as: ； In the formula, is the real value corresponding to the reliability constraint in the kth rule of the ith indicator, is the lower bound of the rule reliability, is the total number of indicators.

5. The device rotation decision method according to claim 2, characterized in that: For the current task, the optimized rotation decision model is used for reasoning to obtain the optimal rotation strategy, which specifically includes: Obtaining the operation duration and the number of operation lines of the current operation task; Determining the activation degree of each rule based on the operation duration and the number of operation lines of the current operation task; activating the language Z-number confidence rules in the optimized rotation decision model based on the activation degree of each rule; when the language Z-number confidence rules are activated, each language Z-number confidence rule includes rule reliability, rule activation weight and rule consequent confidence distribution; Designing an extended proposition evidence reasoning rule, and introducing an extended proposition into the extended proposition evidence reasoning rule; the extended proposition is used to accommodate the unreliability of evidence; Based on the extended proposition evidential reasoning rule with the extended proposition introduced, the discounted confidence is obtained by discounting the confidence distribution of the language Z-number confidence rule after the rule reliability is activated; Adopting an analytical evidence reasoning algorithm to fuse the discounted confidence and obtain a fusion result; Performing anti-discounting processing on the fusion result to obtain the confidence distribution and result reliability of the rotation strategy to be selected; The optimal rotation strategy is obtained based on the confidence distribution of the rotation strategy to be selected and the reliability of the result.

6. The device rotation decision method according to claim 2, characterized in that: Using the historical operation tasks and the optimal rotation strategy corresponding to the historical operation tasks as training data to optimize the rotation decision model specifically includes: Constructing a parameter optimization model of the rotation decision model based on the training data; The particle swarm algorithm is used to search for the optimal solution of the parameter optimization model, and the optimal solution is used as the optimal parameter of the rotation decision model to complete the optimization of the rotation decision model.

7. The device rotation decision method according to claim 2, characterized in that: The parameter optimization model of the rotation decision model is expressed as: In the formula, is the cross entropy objective function, is the discriminant variable, min means taking the minimum value, is the confidence of the nth rotation strategy to be selected, is the operation duration of the task, is the number of rotation strategies to be selected, is the number of operation lines of the operation task, is the number of data sets, is the confidence of the n1th result in the sth data, is the label value of the sth data, Indicates conditions, Represents the discriminant variable of the i-th indicator.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the device rotation decision method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the device rotation decision method described in any one of claims 1 to 7 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the device rotation decision method described in any one of claims 1 to 7 is implemented.

Images