Method for monitoring and evaluating corrosion condition of ball valve

Abstract

The invention discloses a method for monitoring and evaluating the corrosion condition of a ball valve. The method comprises the following steps: acquiring data parameters of a to-be-detected ball valve; acquiring a sampling time point of each monitoring time period according to the sampling frequency coefficient of each monitoring time period; obtaining a corrosion category and a pollution-causing factor obtained at each sampling time point, wherein a sample with the pollution-causing factor exceeding a pollution-causing factor threshold value of the corresponding category is taken as a corrosion sample; a corrosion record is established, the corrosion record is a set of continuous corrosion samples under the same corrosion category, and the corrosion record comprises sampling time points of the corrosion samples and pollution factors of the corrosion samples; according to the method, the corrosion scores of all the corrosion records are obtained, the corrosion condition of the ball valve is evaluated according to the corrosion scores, the corrosion condition is evaluated according to the corrosion scores, and the prediction accuracy is improved.

Description

Technical Field

[0001] This application relates to the field of ball valve corrosion monitoring, specifically to a method for monitoring and evaluating ball valve corrosion. Background Technology

[0002] Ball valves are frequently used in the petroleum, chemical, and water treatment industries. Due to the nature of the media, valves can corrode, such as through chemical corrosion (where the metal surface reacts with the surrounding medium, such as acids, alkalis, and salts), electrochemical corrosion (caused by the formation of an electrolyte film on the metal surface in humid environments), and biological corrosion (where microorganisms grow on the metal surface). All of these can damage the valve's sealing structure, leading to decreased sealing performance and leaks. Leaks caused by valve corrosion can cause environmental pollution. If the leaked fluids are flammable, explosive, or toxic, they can pose serious safety hazards, such as explosions, fires, and harm to human health. In chemical plants, many pieces of equipment are made of metal. Leaks caused by valve corrosion can lead to oil, gas, and seepage, potentially causing explosions, fires, and other major accidents, endangering the environment and property of those in the surrounding area. Therefore, it is necessary to monitor and assess the corrosion damage of ball valves. Currently used methods include sensor technology (such as pressure sensors and flow sensors), visual inspection (camera or laser scanning), acoustic inspection (identifying blockages through sound analysis), vibration analysis, and predictive models based on data analysis. When considering accuracy, real-time performance, cost, maintenance difficulty, and environmental adaptability, sensors may be easily affected by environmental interference, visual inspection is ineffective in the dark or when heavily polluted, acoustic inspection requires high-precision equipment and is costly, and data analysis models require a large amount of data support and may suffer from overfitting. Therefore, it is necessary to find a more reliable and economical detection solution, hoping to integrate multiple technologies to improve the effect. A design that can ensure data accuracy, reduce false alarm rate, and comprehensively optimize the ball valve and its application environment (pipeline) is needed. Summary of the Invention

[0003] To address this problem, this application provides a method for monitoring and evaluating the corrosion of ball valves, comprising the following steps: Obtain the data parameters of the ball valve to be tested. The data parameters include the sampling location of the rust points, the rust type, the value of the contaminating factor corresponding to the rust type, the monitoring period, the basic sampling frequency of the monitoring period, and the sampling frequency coefficient of each monitoring period. The sampling time point for each monitoring period is obtained based on the sampling frequency coefficient for each monitoring period. Obtain the rust category and contaminant at each sampling time point, where the sampling with the contaminant exceeding the contaminant threshold of the corresponding category is considered as rust sampling; Establish a rust record, which is a collection of consecutive rust samples under the same rust category. The rust record includes the sampling time point of the rust sample and the contaminating factor of the rust sample. Obtain the corrosion score for all corrosion records and assess the corrosion condition of the ball valve based on the corrosion score.

[0004] Setting the data parameters includes the following steps: The data parameters include m corrosion categories, where S = [S1, S2, S3, ..., S...]. m ], where the i-th type of corrosion S i The weight is γ i ; The data parameters are set to include n monitoring periods AT, where the monitoring period AT = [AT1, AT2, AT3, ..., AT]. n ], where the j-th monitoring period AT j The sampling frequency coefficient is δ j If the basic sampling frequency is set to H0, then the monitoring period AT will be obtained. j Sampling frequency H j =δ j H0, based on the monitoring period AT j Sampling frequency H j AT period of monitoring was obtained j All sampling time points; During the monitoring period AT j The sampling data obtained at sampling time point ω includes S ω The contaminant for the rust-like category is Q. ω ; The steps for establishing a rust record include: Set S ω The threshold for the contaminating factor in the rust-like category is Q0, and S is obtained based on the sampling time point ω. ω Corrosion-causing agent Q ω When Q ω When Q ≥ 0, this sampling is treated as a corrosion sampling and added to S. ω Corrosion records of rust categories L z In the process, the sampling time point ω is set as the corrosion record L. z The initial sampling time is set to ω, and the pollution factor Q is obtained at the subsequent sampling time point ω+ζ. ω+ζ <Q0, take the sampling time point ω+ζ as the corrosion record L z The end time; The steps for calculating the corrosion fraction of a corrosion record include: Set S ω Corrosion records of rust categories L zIncludes r rust samples, where the peak value of the contaminating factor is Q. max The trough value of the polluting factor is Q. min The pollution growth interval LZ1 includes r1 corrosion samples, the pollution level interval LZ2 includes r2 samples, and the pollution reduction interval includes r3 corrosion samples, where r = r1 + r2 + r3, Q max >Q min ; Based on the sampling time points and contamination factor curves of r rust samples, the rust record L is obtained. z The pollution growth interval LZ1, pollution level interval LZ2, and pollution reduction interval LZ3 are defined. After quantifying the time distribution and corrosion category rise and fall rates of the pollution growth interval LZ1, pollution level interval LZ2, and pollution reduction interval LZ3, the growth interval coefficient ψ1 of the pollution growth interval LZ1, the level interval coefficient ψ2 of the pollution level interval LZ2, and the reduction interval coefficient ψ3 of the pollution reduction interval LZ3 are obtained. Get S ω Corrosion records of rust categories L z Corrosion fraction , where S ω The weight of the corrosion-like category is γ. ω ; The steps to obtain the corrosion score include: Based on the selected preset time period T 预 Set in the preset time period T 预 Given k corrosion records, we get S ω Scores for rust-like categories ; The steps for assessing the corrosion condition of a ball valve include: Set S ω Corrosion category in preset time period T 预 The score threshold F0; when When <F0, determine S ω The type of corrosion did not cause any destructive damage; when When S ≥ F0, determine ω If a rust-like abnormality is detected, a warning will be issued, and the results of the sampling data analysis will be uploaded.

[0005] In the step of establishing the corrosion record, S is obtained based on the sampling time point ω. ω Corrosion-causing agent Q ω When Q ω When Q0 is greater than or equal to 0, the monitoring period AT corresponding to the sampling time point ω will be determined. j sampling frequency coefficient δ j Adjustments were made, and the adjusted monitoring period ATj sampling frequency coefficient δ j '= γ ω δ j , where γ ω For S ω The weight of the corrosion category is 0.5 < γ. ω <1, The value is rounded up to the nearest integer, resulting in the adjusted monitoring period AT. j Sampling frequency H j '=H j δ j '.

[0006] Among them, the sampling frequency is adjusted to H j After that, the pollutant Q was obtained at the next sampling time point ω+1. ω+1 If the polluting factor Q is obtained ω+1 ≥Q ω The monitoring period AT corresponding to the sampling time point ω j sampling frequency coefficient δ j Adjustments are made to the sampling frequency δ. j ''= γ ω δ j ' .

[0007] Among them, the minimum sampling frequency threshold H is set. min When the adjusted sampling frequency reaches the minimum threshold H min At that time, with H min Sampling is performed at the sampling frequency.

[0008] Among them, the rust record L was obtained z The method for determining the pollution growth interval coefficient ψ1 is as follows: Set the duration of the pollution growth interval LZ1 to T1, and the corrosion record L z The total duration is T0, and the time distribution of the pollution growth interval is obtained as λ1 = T1 / T0; Pollution growth rate ε1 = (Q max -Q ω ) / T1; The growth interval coefficient ψ1=λ1ε1 of the pollution growth interval.

[0009] The method for obtaining the equilibrium interval coefficient ψ3 of the pollution reduction interval is as follows: If the duration of the pollution leveling interval is set as T2 and the total duration in the corrosion record is T0, then the time distribution of the pollution growth interval is obtained as λ2 = T2 / T0. The magnitude of the balance ε1=Qω / 2T2; The growth interval coefficient ψ2=λ2ε2 for the pollution growth interval.

[0010] The method for obtaining the pollution reduction coefficient ψ3 in the pollution reduction range is as follows: If the duration of the pollution reduction interval is set to T3 and the total duration in the corrosion record is T0, then the time distribution of the pollution growth interval is obtained as λ3 = T3 / T0. Pollution growth rate ε3 = (Q ω -Q mix ) / T2; The growth interval coefficient ψ3=λ3ε3 for the pollution growth interval.

[0011] This application includes a method that uses historical quality data and a deep learning neural network to predict corrosion sampling, and adjusts the monitoring period settings based on the prediction results. The application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of any of the methods described above.

[0012] Compared with the prior art, the present invention has at least the following beneficial effects: The parameterized settings in this application offer flexibility and adaptability. Different rust spots can cover critical areas, and dynamically adjusting the sampling frequency saves resources while ensuring high-frequency monitoring during critical moments. The sampling frequency can be dynamically adjusted according to actual conditions, increasing sampling frequency during periods of high contamination, thus enabling more timely problem detection, avoiding missed detections, and optimizing resource usage. Recording rust samples and establishing a continuous rust record set allows tracking the duration and trends of rust events, helping to analyze whether the problem is sudden or persistent, which is crucial for subsequent maintenance decisions. Quantitative assessment makes the results more objective, facilitating comparisons of the status of different valves, prioritizing the treatment of valves with severe problems, and potentially predicting remaining lifespan for advance maintenance planning.

[0013] It overcomes the problems that users may face, such as missed detection, waste of resources, or inaccurate assessment. Through parameterization and dynamic adjustment, it improves monitoring efficiency and accuracy, reduces costs, and enables predictive maintenance. By establishing a dynamic parameterized monitoring system, it achieves precise, intelligent, and traceable management of the rust status of ball valves. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0015] Figure 1 A flowchart illustrating the steps for monitoring and evaluating the corrosion of a ball valve as provided in this application embodiment.

[0016] Figure 2 The waveform diagram is a record of the evaluation of the sampling points in the example. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0018] Example 1 See Figure 1 Embodiment 1 of the present invention provides a method for monitoring and evaluating the corrosion of ball valves, comprising the following steps: S1, set data parameters for the pipeline to be inspected. The data parameters include the sampling location of the pipeline to be inspected, the type of corrosion, the threshold of the polluting factor corresponding to the type of corrosion, the monitoring period, the basic sampling frequency of the monitoring period, and the sampling frequency coefficient for each monitoring period. S2, Based on the sampling frequency coefficient of each monitoring period, obtain the sampling time point of each monitoring period; S3, obtain the rust category and contamination factor obtained at each sampling time point, wherein the sampling with the contamination factor exceeding the contamination factor threshold of the corresponding category is regarded as rust sampling; S4, Establish a rust record, which is a collection of consecutive rust samples under the same rust category. The rust record includes the sampling time point of the rust sample and the contaminating factor of the rust sample. S5, calculate the corrosion score of the corrosion record, obtain the corrosion score of all corrosion records within the preset time period, traverse all corrosion points to obtain the corrosion score, and make an evaluation based on the corrosion score.

[0019] Example 2 Based on the same principle, Embodiment 2 of the present invention provides a method for monitoring and evaluating the corrosion of ball valves. This method is a further explanation and refinement of the analysis and processing method in Embodiment 1. The monitoring method for ball valves in Embodiment 2 specifically includes the following steps: In practical applications, selecting appropriate solutions involves categorizing corrosion into solid, liquid, gaseous, and biological types. Monitoring equipment must also employ different detection methods for each of these different corrosion categories. For example, solid particles may require a particle counter, while liquid corrosion necessitates ultrasonic or electrochemical sensors. Special operating conditions, such as high temperature, high pressure, or corrosive environments, must also be considered, placing higher demands on the material selection of monitoring equipment. For instance, monitoring equipment in nuclear power plants or LNG pipelines requires radiation resistance or cryogenic resistance. The installation location and real-time monitoring requirements must also be considered. Space inside ball valves may be limited, necessitating non-invasive or online installation of monitoring equipment to minimize impact on pipeline flow.

[0020] The ball valves in the pipeline are monitored and analyzed based on the properties of the medium in the pipeline and the pipeline flow rate.

[0021] Rust type sampling sensors are installed at appropriate locations within the pipeline to acquire contaminant factors. For example, vibration sensors are installed on the top or bottom flange of the ball valve body to monitor broadband vibrations caused by particles impacting the valve ball. Electromagnetic induction probes are embedded in the valve seat sealing ring groove. A laser particle counter is installed in a straight pipe section 5-10 times the pipe diameter downstream of the ball valve. An ultrasonic thickness sensor is installed in a 45° angled insertion hole on the valve body sidewall to measure the thickness of coking / wax layers. A capacitive liquid film probe is integrated inside the valve ball support shaft to detect viscous liquid residues.

[0022] In this embodiment, the corrosion categories S to be monitored for the pipeline to be inspected include S1 [solid particles], S2 [pipe wall deposits], S3 [corrosive liquids], S4 [harmful gases] and S5 [microorganisms].

[0023] Based on the industrial production situation within the pipeline, the monitoring system is set up with four basic monitoring periods throughout the day, including: the first monitoring period from 6:00 to 12:00, the second monitoring period from 12:00 to 18:00, the third monitoring period from 18:00 to 23:00, and the fourth monitoring period from 23:00 to 6:00.

[0024] In this system, the initial sampling frequency for rust category S1 [solid particles] is set to 120 seconds / time. Sampling frequency coefficients δ are set for the four monitoring periods based on production activities. For example, in this embodiment, the sampling frequency coefficients δ1 and δ2 for the first and second monitoring periods are both 0.5, therefore the sampling frequency H1 for the first monitoring period is 120 seconds / time. 0.5 = 60 seconds / time, the sampling frequency for the second monitoring period is H2 = H1 = 60 seconds / time, the sampling frequency coefficient for the third monitoring period is δ3 = 0.6, therefore the sampling frequency for the third monitoring period is H3 = 120. 0.8 = 96 seconds / time, the sampling frequency coefficient δ4 for the fourth monitoring period is 1.2, therefore the sampling frequency H4 for the fourth monitoring period is 120. 1.2 = 144 seconds / time. In summary, we can obtain all sampling times for the four monitoring periods. Taking the second monitoring period as an example, the sampling frequency coefficient δ2 = 0.5, and the sampling frequency is 60 seconds / time. Therefore, the sampling time points for the second monitoring period are...

[0025] During the second monitoring period, when sampling was conducted at 12:00, the values ​​of the contaminating factors corresponding to all rust categories were obtained. The obtained contaminating factor values ​​were compared with the contaminating factor thresholds of the corresponding rust categories. For example, if the contaminating factor value of solid particles in the three rust categories exceeded the contaminating factor threshold of that category, a rust record L1 was formed. The rust record L1 included the rust category (solid particles), the rust sampling time (12:00), and the contaminating factor. The rust sampling time point (12:00) is taken as the first pollution time of rust record L1 for rust category S1 [solid particles]. The sampling frequency coefficient δ2 of the monitoring period (second monitoring period) corresponding to the first pollution time is adjusted. The adjusted sampling frequency coefficient δ2' of the second monitoring period is = ln 2δ2 , This represents rounding up to the nearest ten. The sampling frequency for the second monitoring period was adjusted to H2'=60. δ2' = 40 seconds / cycle.

[0026] After the sampling frequency is adjusted, the pollution factor Q2 under the solid particulate category will be obtained when sampling is performed at the next sampling time point. When Q2 ≥ Q1, the sampling frequency δ2' of the monitoring period (second monitoring period) corresponding to this sampling time point is adjusted again, and the adjusted sampling frequency δ2'' = ln 2δ2' =30 seconds / sample; After adjusting the sampling frequency, if the pollutant Q3 ≥ Q2 at the next sampling time point, the sampling frequency δ2'' for the corresponding monitoring period (second monitoring period) will be adjusted again, and the adjusted sampling frequency δ2''' = ln 2δ2'' =20 seconds / time; Set the minimum threshold of sampling frequency H0 to 20 seconds / time. Since the sampling frequency after this adjustment has reached the threshold, the sampling frequency will not be adjusted again. Sampling will be performed with H0 as the sampling frequency until the polluting factor no longer increases compared to the previous time.

[0027] When Q0 < Q2 < Q1, continue sampling at the sampling frequency δ2' for this monitoring period; When Q2≤Q0, the sampling time point is taken as the end time point of the corrosion record L1.

[0028] Sampling was performed at the above frequency to obtain rust record L1, which includes 8 rust samples. Based on the rust sampling time point and the contaminating factor, the pollution increase interval, pollution level interval and pollution decrease interval of this solid particle rust record with the change of sampling time point were obtained. Among them, by Figure 2 It can be seen that the current corrosion record includes one pollution increase interval, zero pollution level intervals, and one pollution decrease interval. The pollution increase interval includes four corrosion samples, and the peak value of the contaminating factor is Q. max The pollution reduction range includes four rust sampling points, with the pollution factor trough value being Q. min , where Q max >Q min ; The growth interval coefficient ψ1 of the pollution growth interval, the level interval coefficient ψ2 of the pollution level interval, and the decrease interval coefficient ψ3 of the pollution decrease interval are calculated. The growth interval coefficient ψ1 of the contamination growth interval in the corrosion record is calculated as follows: set the duration of the contamination growth interval to 130 minutes and the total duration in the corrosion record to 90 minutes, then the duration distribution of the contamination growth interval λ1 = 190 / 130 = 0.7 is obtained. The pollution time when the polluting factor reaches its peak during the pollution growth interval is obtained. The duration of the pollution growth interval is 130 seconds. The pollution growth rate ε1 = (Q max -Q1) / 130; The growth interval coefficient ψ1=λ1ε1 for the pollution growth interval; The method for obtaining the equilibrium interval coefficient ψ3 of the pollution reduction interval is as follows: If the duration of the pollution leveling interval is set as T2 and the total duration in the corrosion record is T0, then the time distribution of the pollution growth interval is obtained as λ2 = T2 / T0. The magnitude of the balance ε1=Q ω / 2T2; The growth interval coefficient ψ2=λ2ε2 for the pollution growth interval.

[0029] The method for obtaining the pollution reduction coefficient ψ3 in the pollution reduction range is as follows: If the duration of the pollution reduction interval is set to T3 and the total duration in the corrosion record is T0, then the time distribution of the pollution growth interval is obtained as λ3 = T3 / T0. Pollution growth rate ε3 = (Q ω -Q mix ) / T2; The growth interval coefficient ψ3=λ3ε3 for the pollution growth interval.

[0030] The corrosion fraction F1 = γ1(ψ1ψ2ψ3) = 0.6 in the solid particle corrosion record L1. λ1ε1 λ2ε2 λ3ε3; The score threshold is set to F0. When F1 < F0, it is determined that the pollution did not cause any destructive damage. When F1 ≥ F0, the current corrosion category is judged to have caused a hazardous abnormality, and a warning is issued; Similar sampling points of the same type are obtained as relevant sampling points. By summarizing relevant sampling points, the pollution status and pollution growth of ball valves and their associated pipelines are obtained, a record is formed, the source of pollution is traced, and the historical record is used to form predictions.

[0031] Among them, the sampling location selection and capacity configuration optimization method based on the multi-objective particle swarm optimization algorithm realizes the simulation of spatiotemporal random distribution of supply and demand, and the geospatial analysis prediction and load control method integrates the spatiotemporal changes of meteorological elements for spatiotemporal dynamic matching within the region.

[0032] Based on historical quality data, a deep learning neural network is used to predict pollution occurrences. The monitoring period settings are adjusted according to the prediction results. Specifically, the attribute features sampled from historical data are used as input samples x, and whether a hazardous anomaly occurs is used as output samples y. Set D input attributes, where each attribute is x = [x1; x2; …; x…]. D The corresponding weights are w=[w1; w2;…;w D Let the bias b∈R be set; then the weighted sum z of the input features can be obtained, and the specific formula is: ; Using the ReLU function as the activation function, we have ; In a multilayer feedforward neural network, let The formula for the feedforward neural network to propagate layer by layer through continuous iteration is: ; The composite function is: ; in and This represents the connection weights and biases of all layers in the network. The number of layers in the neural network. For the first The number of neurons in a layer; For the first layer to the first Layer weight matrix; For the first layer to the first Layer bias; For the first The output of layer neurons; Using the cross-entropy loss function, the loss function for sample (x,y) is: ; in, Let y be the one-hot vector representation of y; Given the training set as Each sample The input is fed into the pre-network neural network, and the network output is obtained. Its risk function on the dataset is: ; in, It is a regularization term; λ is a long parameter, the larger λ is, the closer W is to 0: In each iteration of the gradient descent method, the learning rate α is set, and the update methods for parameters W and b are obtained as follows: ; ; Calculate the gradient of the weights and biases of the l-th layer, δ (l) For the error term of the l-th layer: ; ; The iterative formula is obtained as follows: .

[0033] Example 3 Embodiment 3 of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of the monitoring method for ball valves described in Embodiment 1.

[0034] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application. Clearly, those skilled in the art can make various alterations and variations to this application without departing from its spirit and scope. Thus, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

[0035] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in the claims, they should all fall within the protection scope of the present invention.

Claims

1. A method for monitoring and evaluating the corrosion of a ball valve, characterized in that, Includes the following steps: Obtain the data parameters of the ball valve to be tested. The data parameters include the sampling location of the rust points, the rust type, the value of the contaminating factor corresponding to the rust type, the monitoring period, the basic sampling frequency of the monitoring period, and the sampling frequency coefficient of each monitoring period. The sampling time point for each monitoring period is obtained based on the sampling frequency coefficient for each monitoring period. Obtain the rust category and contaminant at each sampling time point, where the sampling with the contaminant exceeding the contaminant threshold of the corresponding category is considered as rust sampling; Establish a rust record, which is a collection of consecutive rust samples under the same rust category. The rust record includes the sampling time point of the rust sample and the contaminating factor of the rust sample. Obtain the corrosion score for all corrosion records and assess the corrosion condition of the ball valve based on the corrosion score.

2. The method for monitoring and evaluating the corrosion of ball valves as described in claim 1, characterized in that, Setting data parameters involves the following steps: The data parameters include m corrosion categories, where S = [S1, S2, S3, ..., S...]. m ], where the i-th type of corrosion S i The weight is γ i ; The data parameters are set to include n monitoring periods AT, where the monitoring period AT = [AT1, AT2, AT3, ..., AT]. n ], where the j-th monitoring period AT j The sampling frequency coefficient is δ j If the basic sampling frequency is set to H0, then the monitoring period AT will be obtained. j Sampling frequency H j =δ j H0, based on the monitoring period AT j Sampling frequency H j AT period was obtained j All sampling time points; During the monitoring period AT j The sampling data obtained at sampling time point ω includes S ω The contaminant for the rust-like category is Q. ω ; The steps for establishing a rust record include: Set S ω The threshold for the contaminating factor in the rust-like category is Q0, and S is obtained based on the sampling time point ω. ω Corrosion-causing agent Q ω When Q ω When Q ≥ 0, this sampling is treated as a corrosion sampling and added to S. ω Corrosion records of rust categories L z In the process, the sampling time point ω is set as the corrosion record L. z The initial sampling time is set to ω, and the pollution factor Q is obtained at the subsequent sampling time point ω+ζ. ω+ζ <Q0, take the sampling time point ω+ζ as the corrosion record L z The end time; The steps for calculating the corrosion fraction of a corrosion record include: Set S ω Corrosion records of rust categories L z Includes r corrosion samples, where the peak value of the contaminating factor is Q. max The trough value of the polluting factor is Q. min The pollution growth interval LZ1 includes r1 corrosion samples, the pollution level interval LZ2 includes r2 samples, and the pollution reduction interval includes r3 corrosion samples, where r = r1 + r2 + r3, Q max >Q min ; Based on the sampling time points and contamination factor curves of r rust samples, the rust record L is obtained. z The pollution growth interval LZ1, pollution level interval LZ2, and pollution reduction interval LZ3 are defined. After quantifying the time distribution and corrosion category rise and fall rates of the pollution growth interval LZ1, pollution level interval LZ2, and pollution reduction interval LZ3, the growth interval coefficient ψ1 of the pollution growth interval LZ1, the level interval coefficient ψ2 of the pollution level interval LZ2, and the reduction interval coefficient ψ3 of the pollution reduction interval LZ3 are obtained. Obtain S ω Corrosion records of rust categories L z Corrosion fraction , where S ω The weight of the corrosion-like category is γ. ω ; The steps to obtain the corrosion score include: Based on the selected preset time period T 预 Set in the preset time period T 预 Given k corrosion records, we get S ω Scores for rust-like categories ; The steps for assessing the corrosion condition of a ball valve include: Set S ω Corrosion category in preset time period T 预 The score threshold F0; when When <F0, determine S ω The type of corrosion did not cause any destructive damage; when When S ≥ F0, determine ω If a rust-like abnormality is detected, a warning will be issued, and the results of the sampling data analysis will be uploaded.

3. The method for monitoring and evaluating the corrosion of ball valves as described in claim 2, characterized in that, In the step of establishing a corrosion record, S is obtained based on the sampling time point ω. ω Corrosion-causing agent Q ω When Q ω When Q0 is greater than or equal to 0, the monitoring period AT corresponding to the sampling time point ω will be determined. j sampling frequency coefficient δ j Adjustments were made, and the adjusted monitoring period AT j sampling frequency coefficient δ j '= γ ω δ j , where γ ω For S ω The weight of the corrosion category is 0.5 < γ. ω <1, The value is rounded up to the nearest integer, resulting in the adjusted monitoring period AT. j Sampling frequency H j '=H j δ j '.

4. The method for monitoring and evaluating the corrosion of ball valves as described in claim 3, characterized in that, The sampling frequency is adjusted to H j After that, the pollutant Q was obtained at the next sampling time point ω+1. ω+1 If the polluting factor Q is obtained ω+1 ≥Q ω The monitoring period AT corresponding to the sampling time point ω j sampling frequency coefficient δ j Adjustments are made, and the adjusted sampling frequency δ j ''= γ ω δ j ' .

5. The method for monitoring and evaluating the corrosion of ball valves as described in claim 4, characterized in that, Set the minimum threshold H for the sampling frequency. min When the adjusted sampling frequency reaches the minimum threshold H min At that time, with H min Sampling is performed at the sampling frequency.

6. The method for monitoring and evaluating the corrosion of ball valves as described in claim 2, characterized in that, Obtain rust record L z The method for determining the pollution growth interval coefficient ψ1 is as follows: Set the duration of the pollution growth interval LZ1 to T1, and the corrosion record L z The total duration is T0, and the time distribution of the pollution growth interval is obtained as λ1 = T1 / T0; Pollution growth rate ε1 = (Q max -Q ω ) / T1; The growth interval coefficient ψ1=λ1ε1 of the pollution growth interval.

7. The method for monitoring and evaluating the corrosion of ball valves as described in claim 2, characterized in that, The method for obtaining the equilibrium interval coefficient ψ3 of the pollution reduction interval is as follows: If the duration of the pollution leveling interval is set as T2 and the total duration in the corrosion record is T0, then the time distribution of the pollution growth interval is obtained as λ2 = T2 / T0. The magnitude of the balance ε1=Q ω / 2T2; The growth interval coefficient ψ2=λ2ε2 for the pollution growth interval.

8. The method for monitoring and evaluating the corrosion of ball valves as described in claim 2, characterized in that, The method for obtaining the pollution reduction coefficient ψ3 in the pollution reduction range is as follows: If the duration of the pollution reduction interval is set to T3 and the total duration in the corrosion record is T0, then the time distribution of the pollution growth interval is obtained as λ3 = T3 / T0. Pollution growth rate ε3 = (Q ω -Q mix ) / T2; The growth interval coefficient ψ3=λ3ε3 for the pollution growth interval.

9. The method for monitoring and evaluating the corrosion of a ball valve as described in claim 8, characterized in that, Based on historical quality data, a deep learning neural network is used to predict rust sampling, and the monitoring period settings are adjusted according to the prediction results.

Images