Evaluation method for easily corroded part of refining device

Abstract

The invention relates to the technical field of petrochemical engineering device monitoring, and particularly discloses a method for evaluating an easily-corroded part of a refining and chemical engineering device, which comprises the following steps of: 1, acquiring corrosion detection data of the easily-corroded part, and collecting historical operation data and an actual working environment of associated equipment; step 2, comprehensively analyzing a corrosion influence factor relationship, evaluating a corrosion state by applying Bass leaf parameter learning, and judging a corrosion risk degree; step 3, predicting a corrosion trend in a preset time period based on a historical data sequence by using the constructed corrosion depth prediction model; according to the method, the causal association of the corrosion process is determined by analyzing the corrosion influence factors and the dependency relationship thereof, the risk probabilities in different corrosion states are evaluated by applying Bass leaf parameter learning, and meanwhile, the corrosion trend in the preset time is predicted by utilizing the constructed corrosion depth prediction model; the high-risk area is identified by integrating the risk probability and the corrosion trend, and evaluation and prevention of parts without obvious corrosion phenomena are achieved.

Description

Technical Field

[0001] The invention belongs to the technical field of petrochemical device monitoring, and in particular relates to an evaluation method for corrosion-prone parts of a refining device. Background Art

[0002] In the process of oil processing, corrosive media such as acids (such as cyclohexane acid) and sulfides in crude oil are bound to exist. They will run through the entire refining production process. The corrosion of the equipment in the refining unit is inevitable, but the severity of the corrosion is different. Therefore, corrosion is the core factor affecting the safe and long-term operation of refining and chemical units. Although the corrosion of refining units is inevitable, the corrosion of refining units can be controlled by reliably monitoring certain key parameters of the corroded parts of the refining units, reliably evaluating these key parameters, and proposing effective and reliable countermeasures.

[0003] There are roughly two types of corrosion monitoring parts in refining and chemical equipment, one is corrosion monitoring at low-temperature corrosion parts, and the other is corrosion monitoring at high-temperature corrosion parts. In refining and chemical enterprises, corrosion monitoring of low-temperature parts is generally based on the uniform corrosion rate detected by corrosion probes such as resistance and inductance, or based on the wall thickness value of the corrosion part measured by the online ultrasonic thickness measuring device. Since the local corrosion reaction process has the characteristics of self-catalysis, the harm of local corrosion reaction is greater than uniform corrosion. If a medium leakage accident occurs due to local corrosion, the consequences of the accident are disastrous. At present, although most domestic petrochemical enterprises have adopted new corrosion monitoring technologies including online monitoring probes or some other monitoring technologies, due to the particularity and complexity of corrosion itself, there is no reliable and scientific evaluation method for the evaluation and selection of monitoring parts.

[0004] Therefore, it is necessary to propose a method for evaluating corrosion-prone parts of a refining device to at least partially solve the problems existing in the prior art.

[0005] The above information disclosed in this background technology is only used to increase the understanding of the background technology of the present invention and therefore, it may include information that does not constitute the prior art known to ordinary technicians in this field. Summary of the invention

[0006] The object of the present invention is to provide a method for evaluating corrosion-prone parts of a refinery to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for evaluating corrosion-prone parts of a refining device, comprising:

[0009] Step 1: Obtain corrosion detection data of corrosion-prone parts, and collect historical operation data and actual working environment of related equipment;

[0010] Step 2: Comprehensively analyze the relationship between corrosion influencing factors, apply Bayesian parameter learning to evaluate the corrosion state, and determine the degree of corrosion risk;

[0011] Step 3: Use the constructed corrosion depth prediction model to predict the corrosion trend within a preset time period based on the historical data sequence;

[0012] Step 4: Identify high-risk areas based on the corrosion risk level and predicted corrosion trend, and conduct follow-up monitoring of high-risk areas;

[0013] Step 5: Develop targeted prevention and maintenance plans, prioritize high-risk areas, and record re-inspection data for systematic analysis;

[0014] Step 6: Organize the entire process data and recommended measures into a written report and update it in a timely manner according to new industry standards and technological advances.

[0015] Preferably, the historical detection data is collected, and the equipment is inspected on site to record external factors that may affect corrosion, understand the corrosion mechanism and influencing factors, determine the corrosion type of the current location, and obtain the design parameters, materials and historical maintenance records of the equipment.

[0016] Preferably, the influencing factors are determined and selected according to the corrosion type, and the influencing factors and their dependencies are sorted out through the knowledge graph. Before learning the Bayesian leaf parameters, the state of each node in the knowledge graph is discretized, the node state representation is divided, and the state probability of each node is initialized.

[0017] Preferably, in the Bayesian leaf parameter learning process, the expected maximum algorithm is introduced to process the parameter estimation of the incomplete data set, and the Bayesian leaf reasoning is performed on the knowledge graph of influencing factors after the node state is initialized through Bayesian leaf parameter learning to obtain the risk probability of the final node being under different corrosion degrees, and the degree of corrosion risk is judged according to the learning results.

[0018] The basic principle of the expectation maximum algorithm (EM) is to assume that the density of the data set D is P(D|θ). If the data is complete, the goal is to maximize the function in formula (1-1):

[0019] L(θ|D)∞p(D|θ) (1-1)

[0020] Where L(θ|D) is the likelihood function of parameter θ.

[0021] However, in engineering practice, some variables may not be observed due to sensor failure, technical limitations or human error, and the data set is usually composed of missing values ​​and is incomplete. If the data is incomplete, D is expressed in formula (1-2) as D = (D obs D unobs ).

[0022] L obs (θ|D obs )∞∫ p (D obs D unobs |θ)dD unobs (1-2)

[0023] The steps of the EM algorithm are as follows:

[0024] ① Expected step: Calculate the expected value of the log-likelihood function based on the initial parameter value or the parameter estimate of the previous iteration:

[0025] E(θ|θ (t) )=∫L obs (D obs ,D unobs )p(D unobs |D obs |θ t )dD unobs

[0026] ②Maximization step: Get the θ that maximizes the expected step length (t+l) value:

[0027]

[0028] Preferably, the corrosion depth prediction model is established by using known data information to determine three smoothing initial values, selecting a smoothing coefficient according to the changing trend of the data, using a genetic algorithm to optimize the three smoothing initial values ​​and the smoothing coefficient, calculating the cubic exponential smoothing value of the historical data for each period, and then calculating the prediction parameters, establishing a cubic exponential smoothing model to predict the corrosion depth, and obtaining a corrosion depth prediction model.

[0029] Assume that the original time series data currently obtained is X={X i |i=1,2,…,N}, where X i is the measured value in the i-th period, and N is the number of samples. The calculation formula of the exponential smoothing value and the prediction parameter after optimization using the genetic algorithm is as follows:

[0030]

[0031] In the formula, are the first, second and third exponential smoothing values ​​of the i-th period, α iis the smoothing coefficient of the i-th period. The smoothing coefficient of each period is set to a different value according to the change of data. i 、b i 、c i are the prediction parameters for the i-th period respectively.

[0032] Preferably, in step four, the risk level is mapped to the equipment using heat map visualization technology, high-risk areas are clearly displayed, real-time data is combined with the prediction model for dynamic monitoring, the assessment of high-risk areas is updated, a data collection and monitoring platform is built, and IoT technology is used to collect and transmit data in real time.

[0033] Preferably, in step five, the prevention and maintenance plan is updated in an event-driven manner, maintenance work is arranged using a scheduling tool, the specific content and results of each maintenance are recorded, the changes in maintenance effects and corrosion risks are analyzed and evaluated, maintenance plans and preventive measures are optimized based on the analysis results, and maintenance and re-inspection reports are generated.

[0034] Preferably, in step six, the assessment process, test results, analysis conclusions and recommendations are organized into a written report, covering all influencing factors and management recommendations, and the assessment results are shared with relevant teams and managers to discuss potential risks and countermeasures.

[0035] A terminal device includes a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, wherein the instructions are suitable for being loaded by the processor and executing any one of the above-mentioned methods for evaluating corrosion-prone parts of a refining device.

[0036] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements the above-mentioned method for evaluating corrosion-prone parts of a refining device.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The present invention analyzes the corrosion influencing factors and their dependencies, clarifies the causal relationship of the corrosion process, applies Bayesian parameter learning to evaluate the risk probability under different corrosion states, and uses the constructed corrosion depth prediction model to predict the corrosion trend within a preset time, and uses a genetic algorithm to determine the optimal smoothing initial value and smoothing coefficient of the model. In this way, corresponding adjustments will be made as the changing trend of the data sequence changes, making the prediction results more accurate, and identifying high-risk areas based on risk probability and corrosion trend, so as to evaluate and prevent areas where no obvious corrosion phenomenon has yet occurred.

[0039] The above summary is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments and features described above, further aspects, embodiments and features of the present invention will be readily apparent by reference to the accompanying drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The present invention is a flow chart of a method for evaluating corrosion-prone parts of a refinery. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] Embodiment 1:

[0043] See also Figure 1 As shown, a method for evaluating corrosion-prone parts of a refinery comprises:

[0044] Step 1: Obtain corrosion detection data of corrosion-prone parts, and collect historical operation data and actual working environment of related equipment;

[0045] Step 2: Comprehensively analyze the relationship between corrosion influencing factors, apply Bayesian parameter learning to evaluate the corrosion state, and determine the degree of corrosion risk;

[0046] Step 3: Use the constructed corrosion depth prediction model to predict the corrosion trend within a preset time period based on the historical data sequence;

[0047] Step 4: Identify high-risk areas based on the corrosion risk level and predicted corrosion trend, and conduct follow-up monitoring of high-risk areas;

[0048] Step 5: Develop targeted prevention and maintenance plans, prioritize high-risk areas, and record re-inspection data for systematic analysis;

[0049] Step 6: Organize the entire process data and recommended measures into a written report and update it in a timely manner according to new industry standards and technological advances.

[0050] Collect historical test data and conduct on-site inspections of equipment to record external factors that may affect corrosion. At the same time, understand the corrosion mechanism and influencing factors, determine the type of corrosion in the current location, and obtain the design parameters, materials and historical maintenance records of the equipment.

[0051] When conducting corrosion inspections, targeted inspection methods are used according to the materials, shapes and environments of different parts, such as ultrasonic testing (UT): used to measure wall thickness; magnetic particle testing (MT): used to detect surface cracks; penetrant testing (PT): used to identify surface defects, etc. Inspection personnel conduct on-site monitoring according to the established inspection plan, and use data recording tools and software to record the data of each inspection point, including temperature, pressure, wall thickness and other information, to ensure the accuracy and traceability of the data.

[0052] The influencing factors are determined according to the corrosion type, and the influencing factors and their dependencies are sorted out through the knowledge graph. Before learning the Bayesian leaf parameters, the state of each node in the knowledge graph is discretized, the node state representation is divided, and the state probability of each node is initialized.

[0053] According to the scope of available data, and with reference to the current specifications and previous literature discretization standards, the nodes are discretized into three states: "low", "medium" and "high", and can be subdivided on this basis as needed. At the same time, some special nodes are presented in two states: "existence" and "non-existence", which describe a state that affirms the cause of a specific node and their interaction with corrosion. The initial state probability of each node is uniformly assigned, and its specific assignment is allocated in combination with the above discretization of the node state.

[0054] In the process of Bayesian leaf parameter learning, the expected maximum algorithm is introduced to process the parameter estimation of incomplete data sets. Through Bayesian leaf parameter learning, Bayesian leaf inference is performed on the knowledge graph of influencing factors after node state initialization to obtain the risk probability of the final node under different degrees of corrosion, and the degree of corrosion risk is judged according to the learning results.

[0055] Establish a corrosion depth prediction model: use known data information to determine three smoothing initial values, select smoothing coefficients according to the changing trend of the data, use genetic algorithms to optimize the three smoothing initial values ​​and smoothing coefficients, calculate the cubic exponential smoothing values ​​of historical data for each period, and then calculate the prediction parameters. Establish a cubic exponential smoothing model for corrosion depth prediction and obtain a corrosion depth prediction model.

[0056] Establish a cubic exponential smoothing model to predict corrosion depth:

[0057]

[0058] Where T is the prediction time interval.

[0059] Before using the genetic algorithm, the parameter encoding method and the appropriate fitness function must be determined first. Since selecting the best initial smoothing value and smoothing coefficient of the model is a multidimensional, high-precision numerical optimization problem, if the traditional binary encoding method is used, a very long encoding sequence will be obtained, which will occupy more memory space and cause large errors. Therefore, the real number encoding method is used, which can not only save the time consumed by encoding and decoding, but also ensure the required accuracy.

[0060] When establishing a cubic exponential smoothing model, it is necessary to seek the minimum value of the loss function (a measure of the difference between the predicted value and the actual value), but the genetic algorithm can only perform maximum optimization, so the original objective function needs to be transformed. Taking the average relative error loss function as an example, the original loss function is shown in formula (2-1).

[0061]

[0062] Where, X i , X i ′ are the measured value and the model predicted value in the i-th period, and N′ is the number of samples used in establishing the model.

[0063] Transform the above formula into a function that measures the fitness value of each individual in the genetic algorithm, formula (2-2) is as follows:

[0064]

[0065] In the formula, the denominator is added with 1 to prevent extreme cases where the denominator becomes 0 and cannot be calculated.

[0066] Embodiment 2:

[0067] See also Figure 1 As shown, this embodiment is basically the same as the previous embodiment, with the difference that in step four, the risk level is mapped to the equipment using heat map visualization technology, high-risk areas are clearly displayed, real-time data is combined with the prediction model for dynamic monitoring, the assessment of high-risk areas is updated, a data collection and monitoring platform is built, and IoT technology is used to collect and transmit data in real time.

[0068] In step five, an event-driven approach is used to update the prevention and maintenance plan. The scheduling tool is used to arrange maintenance work, record the specific content and results of each maintenance, analyze and evaluate the changes in maintenance effects and corrosion risks, optimize the maintenance plan and preventive measures based on the analysis results, and generate maintenance and re-inspection reports.

[0069] In step six, the assessment process, test results, analysis conclusions and recommendations are organized into a written report, covering all influencing factors and management recommendations. The assessment results are shared with relevant teams and managers to discuss potential risks and countermeasures.

[0070] From the above, it can be seen that the present invention analyzes the corrosion influencing factors and their dependencies, clarifies the causal relationship of the corrosion process, applies Bayesian parameter learning to evaluate the risk probability under different corrosion states, and uses the constructed corrosion depth prediction model to predict the corrosion trend within a preset time, and uses a genetic algorithm to determine the optimal smoothing initial value and smoothing coefficient of the model. In this way, corresponding adjustments will be made as the changing trend of the data sequence changes, making the prediction results more accurate, and identifying high-risk areas based on the comprehensive risk probability and corrosion trend, so as to achieve evaluation and prevention of areas where no obvious corrosion phenomenon has yet occurred.

[0071] Embodiment 3:

[0072] An embodiment of the present invention also provides a terminal device, including a processor and a computer-readable storage medium, the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, and the instructions are suitable for being loaded by the processor and executing any one of the above-mentioned methods for evaluating corrosion-prone parts of a refining device.

[0073] The embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, each process of the above-mentioned evaluation method for corrosion-prone parts is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. Among them, the computer-readable storage medium, such as a read-only memory (Read-Only Memory, referred to as ROM), a random access memory (Random ACGess Memory, referred to as RAM), a magnetic disk or an optical disk, etc.

[0074] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0075] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a program, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A way to specify functions in one or more boxes.

[0076] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0077] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0078] In the drawings of the embodiments disclosed in the present invention, only the structures involved in the embodiments disclosed in the present invention are involved, and other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of the present invention can be combined with each other.

[0079] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0080] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating corrosion-prone parts of a refinery, characterized in that: The following steps are involved: Step 1: Obtain corrosion detection data of corrosion-prone parts, and collect historical operation data and actual working environment of related equipment; Step 2: Comprehensively analyze the relationship between corrosion influencing factors, apply Bayesian parameter learning to evaluate the corrosion state, and determine the degree of corrosion risk; Step 3: Use the constructed corrosion depth prediction model to predict the corrosion trend within a preset time period based on the historical data sequence; Step 4: Identify high-risk areas based on the corrosion risk level and predicted corrosion trend, and conduct follow-up monitoring of high-risk areas; Step 5: Develop targeted prevention and maintenance plans, prioritize high-risk areas, and record re-inspection data for systematic analysis; Step 6: Organize the entire process data and recommended measures into a written report and update it in a timely manner according to new industry standards and technological advances.

2. The method for evaluating corrosion-prone parts of a refinery according to claim 1, characterized in that: In step one, historical test data is collected, and the equipment is inspected on site to record external factors that may affect corrosion. At the same time, the corrosion mechanism and influencing factors are understood, the type of corrosion at the current location is determined, and the design parameters, materials, and historical maintenance records of the equipment are obtained.

3. The method for evaluating corrosion-prone parts of a refinery according to claim 2, characterized in that: In step 2, the influencing factors are selected according to the corrosion type, and the influencing factors and their dependencies are sorted out through the knowledge graph. Before learning the Bayesian leaf parameters, the state of each node in the knowledge graph is discretized, the node state representation is divided, and the state probability of each node is initialized.

4. The method for evaluating corrosion-prone parts of a refinery according to claim 3, characterized in that: In the Bayesian leaf parameter learning process, the expected maximum algorithm is introduced to process the parameter estimation of the incomplete data set. The Bayesian leaf reasoning is performed on the knowledge graph of the influencing factors after the node state is initialized through the Bayesian leaf parameter learning to obtain the risk probability of the final node under different corrosion degrees, and the degree of corrosion risk is judged according to the learning results.

5. The method for evaluating corrosion-prone parts of a refinery according to claim 4, characterized in that: The corrosion depth prediction model is established by using known data information to determine three smoothing initial values, selecting smoothing coefficients according to the change trend of the data, using genetic algorithms to optimize the three smoothing initial values ​​and the smoothing coefficients, calculating the cubic exponential smoothing values ​​of the historical data of each period, and then calculating the prediction parameters, establishing a cubic exponential smoothing model to predict the corrosion depth, and obtaining a corrosion depth prediction model.

6. The method for evaluating corrosion-prone parts of a refinery according to claim 5, characterized in that: In step 4, the heat map visualization technology is used to map the risk level to the equipment, clearly display the high-risk areas, combine real-time data with the prediction model for dynamic monitoring, update the assessment of high-risk areas, build a data collection and monitoring platform, and use IoT technology to collect and transmit data in real time.

7. The method for evaluating corrosion-prone parts of a refinery according to claim 6, characterized in that: In step five, an event-driven approach is used to update the prevention and maintenance plan. The scheduling tool is used to arrange maintenance work, record the specific content and results of each maintenance, analyze and evaluate the changes in maintenance effects and corrosion risks, optimize the maintenance plan and preventive measures based on the analysis results, and generate maintenance and re-inspection reports.

8. The method for evaluating corrosion-prone parts of a refinery according to claim 7, characterized in that: In step six, the assessment process, test results, analysis conclusions and recommendations are organized into a written report, covering all influencing factors and management recommendations. The assessment results are shared with relevant teams and managers to discuss potential risks and countermeasures.

9. A terminal device, characterized in that: It includes a processor and a computer-readable storage medium, the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, and the instructions are suitable for being loaded by the processor and executed by a method for evaluating corrosion-prone parts of a refining device as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements a method for evaluating corrosion-prone parts of a refining device as described in claims 1-8.

Images