A constant voltage simulation system for offshore wind power cables

Abstract

The present invention relates to the field of simulation design technology, and specifically discloses a constant voltage simulation system for offshore wind power cables, including: an acquisition module: determining a target segment and a corresponding acquisition period based on an initial value, and drawing a curve of performance scores changing over time within the acquisition period; an analysis module: classifying initial factors; calculating target values, and obtaining working conditions according to the classification of target values ​​and initial factors; recording pending scores, and drawing a curve of pending scores changing over time; and a result module: calculating judgment values, and obtaining target factors in the initial factors based on the judgment values. The present invention significantly improves the accuracy and reliability of simulation evaluation of offshore wind power cable performance.

Description

Technical Field

[0001] The present invention relates to the technical field of simulation design, and in particular to a constant voltage simulation system for offshore wind power cables. Background Art

[0002] Offshore wind power cables are important cable systems specifically used for power transmission in offshore wind farms. Their main function is to transmit the electricity generated by offshore wind turbines from the sea to the land power grid or conversion station, ensuring that the electricity can be transmitted efficiently and safely.

[0003] Cable constant voltage simulation refers to keeping the voltage at both ends of the cable or at specific nodes constant during the simulation process to study the performance and aging characteristics of the cable under long-term fixed voltage conditions. This method can reproduce the combined effects of multiple factors such as electrical, thermal and mechanical factors that the cable may encounter in actual operation in a simulated environment, thereby evaluating the performance of the cable under constant voltage conditions.

[0004] However, although the cable constant voltage simulation method can simulate to a certain extent the complex working environment encountered by the cable in actual operation, the variability of the offshore environment makes it extremely challenging to fully reproduce the coupling effects of all factors. The various electrical, thermal and mechanical stresses interact in the ocean in a complex and varied way, which often exceeds the ability of existing simulation models to accurately capture them. This means that in the process of evaluating cable performance, the simulation results may not cover all possible details and emergencies. Summary of the invention

[0005] The purpose of the present invention is to provide an offshore wind power cable constant voltage simulation system to solve the following technical problems:

[0006] The variability of the offshore environment makes it extremely challenging to fully reproduce the coupled effects of all factors. The ways in which various electrical, thermal and mechanical stresses interact in the ocean are complex and varied, often beyond the ability of existing simulation models to accurately capture.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] An offshore wind power cable constant voltage simulation system includes an acquisition module, an analysis module and a result module, specifically:

[0009] Acquisition module: taking factors affecting cable performance as initial factors, taking the degree of initial factors as initial values, determining target segments on offshore wind power cables based on the initial values, setting acquisition cycles for target segments based on the initial values, acquiring performance parameters of the target segments within the acquisition cycle, wherein the performance parameters represent parameters affecting cable performance, obtaining performance scores of the target segments based on a pre-trained performance scoring model, and drawing a curve f(t) showing performance score changes over time, where t represents time;

[0010] Analysis module: classifies the initial factors, and the number and / or types of initial factors in different classifications are different;

[0011] Calculate the average initial value of the same initial factor in the target segment, take it as the target value total, take the target value as the initial value of the initial factor in the classification, and obtain x working conditions, where x represents the total number of the classification;

[0012] Construct a cable simulation model, record the performance score of the cable simulation model in real time during a preset theoretical period, record it as a pending score, and draw a curve h(t) of the pending score changing with time, wherein the length T of the theoretical period is equal to the acquisition period, and the horizontal coordinate corresponding to the starting point of the theoretical period is 0;

[0013] Result module: adjust the horizontal coordinate corresponding to the starting point of the curve f(t) to 0, obtain the functional relationship F(t) of the curve f(t) at this time, and calculate the judgment value , based on the judgment value, a target factor in the initial factor is obtained, and the target factor is used to perform a constant voltage simulation of an offshore wind power cable.

[0014] As a further solution of the present invention: in the acquisition module, the process of determining the acquisition cycle and the target segment specifically includes:

[0015] Taking any point on the offshore wind power cable as a collection point, a circle with a radius of a preset value r is drawn to obtain a collection circle. Starting from the starting point of the offshore wind power cable, m adjacent collection circles are merged into a collection segment, where m is a preset number, and the initial factors in the collection segment are obtained;

[0016] The acquisition segments are grouped, and the difference value of any two acquisition segments in the same group is less than a preset difference value threshold, and the difference value A=Aac-Abc, where Aac and Abc represent the initial value of the initial factor c in the acquisition segment a and the initial value of the initial factor c in the acquisition segment b, respectively;

[0017] Merge the adjacent acquisition segments in the same group to obtain a target segment, calculate the average initial value of the same initial factor in the target segment, and use it as the target value total;

[0018] The initial value in the target segment is collected in real time as the real-time initial value, and the mean value J of the real-time initial value is calculated. When , the time point at this time is recorded as the cut-off time point T2, CZys represents the preset difference threshold, and the acquisition cycle [T1, T2] is set, and T1 represents the time point for determining the target segment.

[0019] As a further solution of the present invention: in the result module, the process of determining the target factor specifically includes:

[0020] Step S1: obtaining the working condition corresponding to the minimum judgment value, taking it as the target working condition, and taking the initial factors in the target working condition as the factors to be determined;

[0021] Step S2: Generate a pending set djh=(d1, d2, ..., d n ), d n Represent the nth type of pending factor, obtain the intersection of all pending sets, get the target set, take the pending elements in the target set as theoretical factors, calculate the theoretical value P=P1 / Ptot, P1 represents the total number of times the theoretical factor appears in all pending sets, Ptot represents the total number of the pending sets, when the theoretical value P is greater than or equal to the preset theoretical value threshold, take the corresponding theoretical factor as the target factor.

[0022] As a further solution of the present invention: in the acquisition module, the process of obtaining the performance score specifically includes:

[0023] Establishing a database, wherein the database stores performance parameters with annotated performance scores;

[0024] A performance scoring model is established based on the deep learning model, and the performance scoring model is trained and verified through the database; the collected performance parameters are input into the trained and verified performance scoring model, and the performance score is output.

[0025] As a further solution of the present invention: the performance scoring model is trained based on a back-propagation algorithm, and the trained performance scoring model is verified by five-fold cross validation.

[0026] As a further solution of the present invention: the result module further includes the following steps:

[0027] The judgment value Kdy of the working condition corresponding to the target factor is obtained, and the ratio S of the judgment value Kdy≥Kys is counted. When the ratio S≥0.5, it is determined that the target factor is inaccurate, and Kys represents a preset judgment value threshold.

[0028] As a further solution of the present invention: the result module further includes the following steps:

[0029] Step 1: For a single target segment, take the undetermined factor that is not the target factor as the first factor, and sort the first factor in descending order according to the size of the corresponding target value in the target segment Y to obtain a first sort;

[0030] Obtain a reference working condition, wherein the initial factors in the reference working condition are the first factor y1 and the target factor, wherein the first factor y1 is the first factor in the first ranking, and obtain a judgment value K1 of the reference working condition, and when the judgment value K1<Kys, use the first factor y1 as the second factor;

[0031] Remove the first factor y1 from the first sorting, and repeat the above steps until the judgment value corresponding to a first factor is greater than or equal to Kys, to obtain a second factor of a single target segment;

[0032] Obtaining the second factors corresponding to all target segments, taking the second factors as pending factors, and executing step S2 to obtain new target factors;

[0033] Step 2: Determine whether the new target factor is accurate. If the new target factor is inaccurate, execute step 1 again. After repeating step 1, determine whether the new target factor is accurate. If the corresponding target factor is still inaccurate after repeating step 1 for the Nth time, execute step 3.

[0034] Step 3: Set the new judgment value threshold , ΔK represents the preset judgment value update threshold, and it is judged again whether the initially determined target factor is accurate.

[0035] Beneficial effects of the present invention: In this solution, first, the performance data of the cable during actual operation is collected. These data serve as the benchmark for subsequent analysis and provide a solid reference for comparing simulation results. By obtaining key data reflecting the impact of the real environment, it helps to focus on the factors that really affect the performance of the cable in subsequent analysis, avoids simulating all details in the simulation, and reduces the pressure on the demand for comprehensive reproduction of complex environments. It is worth noting that the initial factors include but are not limited to ambient temperature, seawater salinity, etc. Taking ambient temperature as an example, the initial value reflects the ambient temperature. The higher the ambient temperature, the higher the initial value. The specific corresponding relationship can be calibrated by experiment. Afterwards, by using different working conditions The performance scoring curve during the acquisition simulation process is aimed at decomposing complex environmental factors into multiple specific working conditions and establishing corresponding performance change trends for each working condition. It can be understood that different target segments correspond to different working conditions, one working condition corresponds to a classification, and the initial value of the initial factor in the classification corresponding to the working condition is the target value. Finally, by comparing the actual acquisition data with the simulation data curve, the purpose is to determine the target factor that best reflects the actual situation in a complex simulation environment. This step not only simplifies the analysis process, but also avoids in-depth simulation of all details by focusing on these key factors, greatly reducing the reliance on fully reproducing complex coupling effects, while improving the efficiency and accuracy of the final evaluation and optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The present invention will be further described below in conjunction with the accompanying drawings.

[0037] Figure 1 It is a structural schematic diagram of a constant voltage simulation system for offshore wind power cables of the present invention. DETAILED DESCRIPTION

[0038] 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.

[0039] See also Figure 1 As shown, the present invention is an offshore wind power cable constant voltage simulation system, including a collection module, an analysis module and a result module, specifically:

[0040] Acquisition module: taking factors affecting cable performance as initial factors, taking the degree of initial factors as initial values, determining target segments on offshore wind power cables based on the initial values, setting acquisition cycles for target segments based on the initial values, acquiring performance parameters of the target segments within the acquisition cycle, wherein the performance parameters represent parameters affecting cable performance, obtaining performance scores of the target segments based on a pre-trained performance scoring model, and drawing a curve f(t) showing performance score changes over time, where t represents time;

[0041] Analysis module: classifies the initial factors, and the number and / or types of initial factors in different classifications are different;

[0042] Calculate the average initial value of the same initial factor in the target segment, take it as the target value total, take the target value as the initial value of the initial factor in the classification, and obtain x working conditions, where x represents the total number of the classification;

[0043] Construct a cable simulation model, record the performance score of the cable simulation model in real time during a preset theoretical period, record it as a pending score, and draw a curve h(t) of the pending score changing with time, wherein the length T of the theoretical period is equal to the acquisition period, and the horizontal coordinate corresponding to the starting point of the theoretical period is 0;

[0044] Result module: adjust the horizontal coordinate corresponding to the starting point of the curve f(t) to 0, obtain the functional relationship F(t) of the curve f(t) at this time, and calculate the judgment value , based on the judgment value, a target factor in the initial factor is obtained, and the target factor is used to perform a constant voltage simulation of an offshore wind power cable.

[0045] It should be noted that the performance data of the cable during actual operation is collected. These data serve as the benchmark for subsequent analysis and provide a solid reference for comparing simulation results. By obtaining key data reflecting the impact of the real environment, it helps to focus on the factors that really affect the performance of the cable in subsequent analysis, avoid simulating all details in the simulation, and reduce the pressure on the comprehensive reproduction of complex environments. It is worth noting that the initial factors include but are not limited to ambient temperature, seawater salinity, etc. Taking ambient temperature as an example, the initial value reflects the ambient temperature. The higher the ambient temperature, the higher the initial value. The specific corresponding relationship can be calibrated by experiment. Afterwards, the performance scoring curves in the simulation process are collected under different working conditions. The purpose is to The complex environmental factors are decomposed into a variety of specific working conditions, and the corresponding performance change trend is established for each working condition. Through these curves, the specific impact of each working condition on the cable performance can be clearly observed; it can be understood that different target sections correspond to different working conditions, one working condition corresponds to a classification, and the initial value of the initial factor in the classification corresponding to the working condition is the target value; finally, by comparing the actual collected data with the simulation data curve, the purpose is to determine the target factor that best reflects the actual situation in a complex simulation environment. This step not only simplifies the analysis process, but also avoids in-depth simulation of all details by focusing on these key factors, greatly reducing the reliance on comprehensive reproduction of complex coupling effects, while improving the efficiency and accuracy of the final evaluation and optimization.

[0046] In another preferred embodiment of the present invention, in the acquisition module, the process of determining the acquisition period and the target segment specifically includes:

[0047] Taking any point on the offshore wind power cable as a collection point, a circle with a radius of a preset value r is drawn to obtain a collection circle. Starting from the starting point of the offshore wind power cable, m adjacent collection circles are merged into a collection segment, where m is a preset number, and the initial factors in the collection segment are obtained;

[0048] The acquisition segments are grouped, and the difference value of any two acquisition segments in the same group is less than a preset difference value threshold, and the difference value A=Aac-Abc, where Aac and Abc represent the initial value of the initial factor c in the acquisition segment a and the initial value of the initial factor c in the acquisition segment b, respectively;

[0049] Merge the adjacent acquisition segments in the same group to obtain a target segment, calculate the average initial value of the same initial factor in the target segment, and use it as the target value total;

[0050] The initial value in the target segment is collected in real time as the real-time initial value, and the mean value J of the real-time initial value is calculated. When , the time point at this time is recorded as the cut-off time point T2, CZys represents the preset difference threshold, and the acquisition cycle [T1, T2] is set, and T1 represents the time point for determining the target segment.

[0051] In another preferred embodiment of the present invention, in the result module, the process of determining the target factor specifically includes:

[0052] Step S1: obtaining the working condition corresponding to the minimum judgment value, taking it as the target working condition, and taking the initial factors in the target working condition as the factors to be determined;

[0053] Step S2: Generate a pending set djh=(d1, d2, ..., d n ), d n Represent the nth type of pending factor, obtain the intersection of all pending sets, get the target set, take the pending elements in the target set as theoretical factors, calculate the theoretical value P=P1 / Ptot, P1 represents the total number of times the theoretical factor appears in all pending sets, Ptot represents the total number of the pending sets, when the theoretical value P is greater than or equal to the preset theoretical value threshold, take the corresponding theoretical factor as the target factor.

[0054] In another preferred embodiment of the present invention, in the acquisition module, the process of obtaining the performance score specifically includes:

[0055] Establishing a database, wherein the database stores performance parameters with annotated performance scores;

[0056] A performance scoring model is established based on the deep learning model, and the performance scoring model is trained and verified through the database; the collected performance parameters are input into the trained and verified performance scoring model, and the performance score is output.

[0057] It should be noted that first, a database containing performance parameters with annotated performance scores was established, which provided a rich and accurate data source for subsequent model training, thereby ensuring that the features learned by the model are representative and effective. At the same time, the construction of the database reduced the duplication of data acquisition and human errors. Next, a performance scoring model was established based on the deep learning model, and the model was fully trained and verified using the aforementioned database. The benefit of this step is that with the help of advanced algorithms and a large amount of data, the model can automatically discover the complex patterns and laws that affect cable performance, with the aim of improving the accuracy and reliability of performance scoring, thereby helping engineers to more scientifically evaluate cable conditions. Finally, the performance parameters collected in real time are input into the trained and verified model, and the performance score is output. The advantage of this integrated operation is that it realizes automated and intelligent performance evaluation. Performance parameters include but are not limited to insulation resistance: reflecting the integrity and leakage current of the cable insulation layer. High insulation resistance usually indicates good insulation condition. Partial discharge level: monitoring micro-discharge phenomena inside or on the surface of the insulating material. Continuously increasing partial discharge may indicate insulation aging or potential failure.

[0058] In another preferred embodiment of the present invention, the performance scoring model is trained based on a back propagation algorithm, and the trained performance scoring model is verified by five-fold cross validation.

[0059] It should be noted that the back propagation algorithm is a commonly used supervised learning algorithm. By calculating the gradient of the output error relative to the weights of each layer, the network weights are updated backward layer by layer, so that the model can continuously adjust the parameters to minimize the prediction error, thereby improving the accuracy of the performance score. The five-fold cross-validation divides the data set into five parts, trains the model with four parts in turn, and verifies the remaining part, and repeats five times to obtain the average performance index. This method can effectively prevent overfitting and ensure the generalization ability of the model on unseen data. Combining the efficient weight optimization of the back propagation algorithm and the robust evaluation of the five-fold cross-validation, the purpose of this scheme is to establish a performance scoring model to make the model training more accurate and reliable, so that the cable performance score can be accurately output in practical applications.

[0060] In another preferred embodiment of the present invention, the result module further includes the following steps:

[0061] The judgment value Kdy of the working condition corresponding to the target factor is obtained, and the ratio S of the judgment value Kdy≥Kys is counted. When the ratio S≥0.5, it is determined that the target factor is inaccurate, and Kys represents a preset judgment value threshold.

[0062] It can be understood that by quantitatively evaluating the performance of the target factor under different working conditions, it is possible to test whether the previously determined target factor has consistent influence and accuracy, so as to timely discover and correct possible misjudgments. In specific operations, the ratio S of the statistical judgment value greater than or equal to a certain credible threshold Kys can objectively measure the proportion of the target factor in the expected performance. The ratio S reflects the reliability of the model for the factor. If the statistically derived ratio S reaches or exceeds 50%, it indicates that in more than half of the cases, the factor fails to effectively predict or explain the cable performance, and thus the target factor is judged to be inaccurate. It should be noted that the working condition corresponding to the target factor, that is, the initial factor is the target factor and the number of initial factors is equal to the working condition of the target factor.

[0063] In another preferred embodiment of the present invention, the result module further includes the following steps:

[0064] Step 1: For a single target segment, take the undetermined factor that is not the target factor as the first factor, and sort the first factor in descending order according to the size of the corresponding target value in the target segment Y to obtain a first sort;

[0065] Obtain a reference working condition, wherein the initial factors in the reference working condition are the first factor y1 and the target factor, wherein the first factor y1 is the first factor in the first ranking, and obtain a judgment value K1 of the reference working condition, and when the judgment value K1<Kys, use the first factor y1 as the second factor;

[0066] Remove the first factor y1 from the first sorting, and repeat the above steps until the judgment value corresponding to a first factor is greater than or equal to Kys, to obtain a second factor of a single target segment;

[0067] Obtaining the second factors corresponding to all target segments, taking the second factors as pending factors, and executing step S2 to obtain new target factors;

[0068] Step 2: Determine whether the new target factor is accurate. If the new target factor is inaccurate, execute step 1 again. After repeating step 1, determine whether the new target factor is accurate. If the corresponding target factor is still inaccurate after repeating step 1 for the Nth time, execute step 3.

[0069] Step 3: Set the new judgment value threshold , ΔK represents the preset judgment value update threshold, and it is judged again whether the initially determined target factor is accurate.

[0070] It can be understood that the initially determined target factor is the target factor in the initial factors obtained for the first time based on the judgment value.

[0071] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An offshore wind power cable constant voltage simulation system, characterized in that: It includes acquisition module, analysis module and result module, specifically: Acquisition module: taking factors affecting cable performance as initial factors, taking the degree of initial factors as initial values, determining target segments on offshore wind power cables based on the initial values, setting acquisition cycles for target segments based on the initial values, acquiring performance parameters of the target segments within the acquisition cycle, wherein the performance parameters represent parameters affecting cable performance, obtaining performance scores of the target segments based on a pre-trained performance scoring model, and drawing a curve f(t) showing performance score changes over time, where t represents time; Analysis module: classifies the initial factors, and the number and / or types of initial factors in different classifications are different; Calculate the average initial value of the same initial factor in the target segment, take it as the target value total, take the target value as the initial value of the initial factor in the classification, and obtain x working conditions, where x represents the total number of the classification; Construct a cable simulation model, record the performance score of the cable simulation model in real time during a preset theoretical period, record it as a pending score, and draw a curve h(t) of the pending score changing with time, wherein the length T of the theoretical period is equal to the acquisition period, and the horizontal coordinate corresponding to the starting point of the theoretical period is 0; Result module: adjust the horizontal coordinate corresponding to the starting point of the curve f(t) to 0, obtain the functional relationship F(t) of the curve f(t) at this time, and calculate the judgment value , based on the judgment value, a target factor in the initial factor is obtained, and the target factor is used to perform a constant voltage simulation of an offshore wind power cable.

2. An offshore wind power cable constant voltage simulation system according to claim 1, characterized in that: In the acquisition module, the process of determining the acquisition period and the target segment specifically includes: Taking any point on the offshore wind power cable as a collection point, a circle with a radius of a preset value r is drawn to obtain a collection circle. Starting from the starting point of the offshore wind power cable, m adjacent collection circles are merged into a collection segment, where m is a preset number, and the initial factors in the collection segment are obtained; The acquisition segments are grouped, and the difference value of any two acquisition segments in the same group is less than a preset difference value threshold, and the difference value A=Aac-Abc, where Aac and Abc represent the initial value of the initial factor c in the acquisition segment a and the initial value of the initial factor c in the acquisition segment b, respectively; Merge the adjacent acquisition segments in the same group to obtain a target segment, calculate the average initial value of the same initial factor in the target segment, and use it as the target value total; The initial value in the target segment is collected in real time as the real-time initial value, and the mean value J of the real-time initial value is calculated. When , the time point at this time is recorded as the cut-off time point T2, CZys represents the preset difference threshold, and the acquisition cycle [T1, T2] is set, and T1 represents the time point for determining the target segment.

3. The offshore wind power cable constant voltage simulation system according to claim 1 is characterized in that: In the result module, the process of determining the target factor specifically includes: Step S1: obtaining the working condition corresponding to the minimum judgment value, taking it as the target working condition, and taking the initial factors in the target working condition as the factors to be determined; Step S2: Generate a pending set djh=(d1, d2, ..., d n ), d n Represent the nth type of pending factor, obtain the intersection of all pending sets, get the target set, take the pending elements in the target set as theoretical factors, calculate the theoretical value P=P1 / Ptot, P1 represents the total number of times the theoretical factor appears in all pending sets, Ptot represents the total number of the pending sets, when the theoretical value P is greater than or equal to the preset theoretical value threshold, take the corresponding theoretical factor as the target factor.

4. The offshore wind power cable constant voltage simulation system according to claim 1 is characterized in that: In the acquisition module, the process of obtaining performance scores specifically includes: Establishing a database, wherein the database stores performance parameters with annotated performance scores; A performance scoring model is established based on the deep learning model, and the performance scoring model is trained and verified through the database; the collected performance parameters are input into the trained and verified performance scoring model, and the performance score is output.

5. An offshore wind power cable constant voltage simulation system according to claim 4, characterized in that: The performance scoring model is trained based on a back-propagation algorithm, and the trained performance scoring model is verified through five-fold cross validation.

6. The offshore wind power cable constant voltage simulation system according to claim 3 is characterized in that: The result module also includes the following steps: The judgment value Kdy of the working condition corresponding to the target factor is obtained, and the ratio S of the judgment value Kdy≥Kys is counted. When the ratio S≥0.5, it is determined that the target factor is inaccurate, and Kys represents a preset judgment value threshold.

7. An offshore wind power cable constant voltage simulation system according to claim 6, characterized in that: The result module also includes the following steps: Step 1: For a single target segment, take the undetermined factor that is not the target factor as the first factor, and sort the first factor in descending order according to the size of the corresponding target value in the target segment Y to obtain a first sort; Obtain a reference working condition, wherein the initial factors in the reference working condition are the first factor y1 and the target factor, wherein the first factor y1 is the first factor in the first ranking, and obtain a judgment value K1 of the reference working condition, and when the judgment value K1<Kys, use the first factor y1 as the second factor; Remove the first factor y1 from the first sorting, and repeat the above steps until the judgment value corresponding to a first factor is greater than or equal to Kys, to obtain a second factor of a single target segment; Obtaining the second factors corresponding to all target segments, taking the second factors as pending factors, and executing step S2 to obtain new target factors; Step 2: Determine whether the new target factor is accurate. If the new target factor is inaccurate, execute step 1 again. After repeating step 1, determine whether the new target factor is accurate. If the corresponding target factor is still inaccurate after repeating step 1 for the Nth time, execute step 3. Step 3: Set the new judgment value threshold , ΔK represents the preset judgment value update threshold, and it is judged again whether the initially determined target factor is accurate.

Images