Control method of flood control turnover gate

Abstract

The invention discloses a control method of a flood control flip gate, which comprises the following steps of: decomposing a problem into a target layer, a criterion layer, a sub-criterion layer and a scheme layer, constructing a judgment matrix through data analysis, and calculating a characteristic value and a characteristic vector to determine the weight of each factor; comprehensive evaluation is carried out, scores are assigned to different schemes according to the water level condition, and weighted scores are calculated to determine the height of the flood gate; and finally, data correction and scheme division are carried out, the lifting height of the flood gate is divided into four grades, the gate is lifted to the corresponding height according to the grades, and reasonable adjustment of the flood gate under different flood risks is ensured.

Description

Technical Field

[0001] The present invention relates to the fields of urban transportation and intelligent control technology, and particularly relates to a control method for a flood prevention flipping gate. Background Art

[0002] As an important part of urban transportation, the subway bears a huge passenger flow. Under extreme weather conditions, such as floods caused by heavy rain, the safety of passengers may be threatened. Therefore, the implementation of subway flood prevention measures is crucial for protecting the lives and safety of passengers.

[0003] The traditional subway waterproofing methods mainly include the following:

[0004] (1) Sandbags: Simple in structure and low in cost, but completely dependent on manual operation, with low efficiency;

[0005] (2) Side-opening and closing flood prevention gates: Increased flexibility compared to traditional methods, but insufficient sealing performance, with a risk of water leakage.

[0006] (3) Power flipping flood prevention gates: In the past, for traditional flipping gates, the height of the gate was fixed, and its flood prevention ability was rather limited. Whether the water level rose or fell, the gate could only be maintained at a fixed and non-adjustable height, which undoubtedly limited its flexibility in dealing with complex flood situations.

[0007] (4) Vertically lifting flood prevention gates: Require deep excavation at the subway entrance, increasing the construction difficulty and bringing many inconveniences to later maintenance. There are also problems such as high labor costs and complex installation hidden behind it.

[0008] (5) Hydraulically driven flipping flood prevention gates: Sturdy and durable, with a wider excavation surface and shallower depth, but the characteristic that the height of its gate is fixed limits the further improvement of its flood prevention ability.

[0009] Therefore, based on the existing gate design scheme, it is necessary to overcome the many deficiencies of the original flood prevention gates and provide a new and more efficient solution for urban flood prevention safety. Summary of the Invention

[0010] In view of this, the purpose of the present invention is to provide a control method for a flood prevention flipping gate, which improves the flood prevention method of the existing hydraulically driven flood prevention flipping gate, realizes the function of dynamically adjusting the gate height, enhances the intelligent level of the flood prevention gate, improves its flood prevention effect, and provides a more solid guarantee for subway flood prevention safety.

[0011] The purpose of the present invention is achieved through the following technical solutions:

[0012] A control method for a flood control flip gate of the present invention includes the automatic control of the gate rising, that is, by collecting air humidity, precipitation, green space density, and water level data as factor factors, jointly analyzing their respective weights, obtaining the water level weight and the partition interval weight, comprehensively analyzing and setting different levels of gate rising schemes for the rising height of the flood control gate, and determining the gate rising scheme and transmitting the gate flip rising signal according to the water level rising data of the real-time collected water level trigger, so as to control the gate rising.

[0013] Further, the method also includes the automatic control of the gate descending, that is, controlling the gate to descend according to the water level descending data of the real-time collected water level trigger.

[0014] Further, the automatic control of the gate rising specifically includes the following steps:

[0015] Step S1: Decompose the problem of adjusting the height of the flood control gate into multiple levels, including the target layer, criterion layer, sub-criterion layer, and scheme layer, and construct a judgment matrix:

[0016] Step S2: By calculating the eigenvalues and eigenvectors of the judgment matrix, the relative weight of each factor can be obtained;

[0017] Step S3: According to the weight of each factor and the score of the evaluation object, calculate the comprehensive evaluation score, and calculate the height of the flood control gate determined under different water level schemes;

[0018] Step S4: Divide into different levels of gate rising schemes according to the comprehensive evaluation score.

[0019] Further, in step S3, before calculating the comprehensive evaluation score, obtain the importance of each factor by calculating the information entropy and information gain of the data set and normalize it; specifically include the following sub-steps:

[0020] Step S31: First calculate the total information entropy, and the calculation formula is:

[0021]

[0022] Among them, pi represents the sample proportion of category i, that is, the number of samples in category i divided by the total number of samples; n represents the number of label types; the calculation result will obtain the overall entropy value of the data before partitioning;

[0023] Step S32: Calculate the information gain:

[0024] Traverse each feature f and calculate all possible values V(f) of the feature;

[0025] For each value v ∈ V(f), divide the data set S into sub-data sets Sv according to the value v of the feature f;

[0026] Calculate the entropy H(Sv) of each sub-dataset;

[0027] Calculate the weighted average entropy after splitting. The calculation formula is:

[0028]

[0029] Calculate the information gain. The calculation formula is:

[0030] IG(S, f) = H(S) - weighted average entropy;

[0031] Step S33: Select the optimal splitting feature:

[0032] Compare the information gains of all features and select the feature with the largest gain as the splitting feature of the current node;

[0033] Step S34: Recursive splitting:

[0034] Recursively call the above steps for each value of the selected splitting feature to continue constructing child nodes until all features are used up;

[0035] Step S35: Output the importance of features. The feature importance measures the contribution degree of each feature in the model, mainly based on the reduction of impurity when splitting nodes;

[0036]

[0037] Step S36: Output the result after normalizing the importance.

[0038] Furthermore, in the above step S1, the target layer is divided into the height of the dynamic adjustment of the flood control gate; the criterion layer is divided into air humidity, precipitation, green space density, water level; the sub-criterion layer is divided into air humidity and precipitation. Air humidity includes relative humidity, absolute humidity, humidity change rate; precipitation includes rainfall, precipitation intensity, precipitation frequency; the scheme layer is divided into schemes of different rising heights of the flood control gate, including four levels: low, medium, high, and extremely high.

[0039] Furthermore, in the above step S1, to construct the judgment matrix for each level of elements, the judgment matrix is constructed by pairwise comparison. The elements in the judgment matrix represent the relative importance between two factors. It is stipulated that the weight of each factor relative to the previous factor is 1, indicating equal importance; or greater than 1, indicating more important; among them, the green space density is quantified according to the normalized difference vegetation index. The calculation formula is:

[0040]

[0041] Among them, NDVI is the normalized difference vegetation index, NIR is the reflection value of the near-infrared band, and R is the reflection value of the red light band.

[0042] Furthermore, in the step S2, the importance of each factor in the risk assessment is reflected by weights, and the specific steps include:

[0043] Find the eigenvector of the judgment matrix. The eigenvector is an attribute of the matrix and satisfies the following equation:

[0044] Ax = cx;

[0045] where A is the judgment matrix, c is the eigenvalue, and x is the eigenvector;

[0046] After finding the eigenvector, normalize it so that the sum of its elements is 1. The normalized eigenvector is the relative weight of each factor.

[0047] Furthermore, in the step S4, according to the weight of each factor and the score of the evaluation object, calculate the comprehensive evaluation score, which reflects the overall performance of the evaluation object in the risk assessment;

[0048] First, conduct data analysis based on the criterion of water level to determine the scores of different schemes, where the score of the low water level is the lowest and the score of the extremely high water level is the highest;

[0049] Then, multiply the score of each scheme by the weight of the corresponding criterion. The formula for calculating the weighted score is as follows:

[0050]

[0051] where C represents the weighted score; W represents the weight; I represents the importance calculated by the decision tree; S represents the scheme score;

[0052] Finally, add the flood control gate height index. Determine a benchmark value based on the analysis of historical flood data, and determine the adjustment of the flood control gate height under different water level schemes. The formula for calculating the flood control gate height is as follows:

[0053] H = C × Standard;

[0054] where H represents the flood control gate height determined under different water level schemes, C represents the weighted score, and Standard represents the flood control gate height benchmark, and calculate the flood control gate height determined under different water level schemes.

[0055] Furthermore, in the step S3, according to the flood control gate height determined under different water level schemes obtained by calculation, correct the flood control gate height and divide it into four different grades of gate rising schemes: low, medium, high, and extremely high.

[0056] Furthermore, the method also includes a manual control mode for the gate, and the manual control mode is manually controlled by humans to independently select the rising and falling heights of the gate at each stage.

[0057] The beneficial effects of the present invention are as follows:

[0058] (1) When the water level is low, the flipping height of the gate in the method of the present invention is low. Compared with the fixed flipping height design of the traditional gate device, the present invention dynamically adjusts the height by improving the device, avoiding problems such as great difficulty in rescuing citizens such as the elderly, the weak, the sick, the disabled, and pregnant women, and improving the efficiency of personnel for rescue or refuge;

[0059] (2) The present invention includes automatic lifting and manual intervention modes, with flexibility, making it more in line with the actual flood control requirements and more reflecting the humanized design concept;

[0060] (3) The present invention adopts a lifting design. According to the inconsistent heights of each layer of water pressure, the height and thickness are dynamically changed, thus saving material costs;

[0061] (4) The device utilized by the present invention occupies a small area, and subsequent maintenance and repair become easier. There is no need to deeply excavate the ground, having advantages such as low labor costs and maintenance costs.

[0062] Other advantages, objectives, and features of the present invention will be described to some extent in the subsequent description, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the following description and the foregoing claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the drawings, where:

[0064] Figure 1 is the overall functional design flow chart of Embodiment 1 of the present invention;

[0065] Figure 2 is the dynamic change state flow chart of the lifting gate at the subway entrance;

[0066] Figure 3 is the overall functional design flow chart of Embodiment 2 of the present invention;

[0067] Figure 4 is the flow chart of calculating weights by the decision tree algorithm according to the information entropy and information gain theory. DETAILED DESCRIPTION OF THE INVENTION

[0068] The following will refer to the drawings to describe the preferred embodiments of the present invention in detail. It should be understood that the preferred embodiments are only for illustrating the present invention, rather than for limiting the protection scope of the present invention.

[0069] A control method for a flood control flip gate of the present invention includes an automatic control mode, and the automatic control mode includes the rising automatic control and the falling automatic control of the gate;

[0070] Generally speaking, the rising automatic control jointly analyzes the respective weights by collecting air humidity, precipitation, green space density, and water level data as factor factors, obtains the water level weight and the partition interval weight, and determines the gate rising schemes at different levels of the rising height of the flood control gate through comprehensive analysis; in practice, according to the water level rising data of the water level trigger collected in real time, the gate rising scheme is determined and the gate flip rising signal is transmitted, thereby controlling the rising of the gate.

[0071] The falling automatic control of the gate is to control the falling of the gate through the water level falling data of the water level trigger collected in real time.

[0072] The rising automatic control specifically includes the following steps:

[0073] Step S1: Decompose the problem of adjusting the height of the flood control gate into multiple levels, including the target layer, the criterion layer, the sub-criterion layer, and the scheme layer, and construct a judgment matrix:

[0074] Step S2: By calculating the eigenvalues and eigenvectors of the judgment matrix, the relative weight of each factor can be obtained;

[0075] Step S3: According to the weight of each factor and the score of the evaluation object, calculate the comprehensive evaluation score, and calculate the height of the flood control gate determined under different water level schemes;

[0076] Step S4: Divide the gate rising schemes into different levels according to the comprehensive evaluation score.

[0077] Embodiment 1

[0078] The above automatic control mode is implemented by the Analytic Hierarchy Process (AHP). The overall function design flowchart of this embodiment is as Figure 1 shown. The following will further elaborate on the specific schemes of each step. The process in Step S1 includes:

[0079] 1. Decompose the problem of adjusting the height of the flood control gate into multiple levels, including the target layer, the criterion layer, the sub-criterion layer, and the scheme layer. Specifically, it includes:

[0080] (1) The target layer is divided into the height of the dynamic adjustment of the flood control gate;

[0081] (2) The criterion layer is divided into air humidity, precipitation, green space density, and water level;

[0082] (3) The sub-criterion layer is divided into air humidity and precipitation. Air humidity includes relative humidity, absolute humidity, and humidity change rate; precipitation includes rainfall, precipitation intensity, and precipitation frequency. The sub-criterion layer plays a role in further refining the decision-making structure, mainly used in complex multi-criteria decision-making problems to decompose the problem into more detailed levels, thereby helping to more accurately express the relationships between various aspects and factors of the decision-making problem, specifically reflected in aspects such as refining the main criterion, clarifying influencing factors, improving the accuracy of weight allocation, and enhancing the model's interpretability.

[0083] (4) The solution layer is divided into solutions with different rising heights of flood control gates, including four levels: low, medium, high, and extremely high.

[0084] 2. Constructing the judgment matrix is to construct the judgment matrix for each level of elements through pairwise comparison. The elements in the judgment matrix represent the relative importance between two factors. It is stipulated that the weight of each factor relative to the previous factor is 1, indicating equal importance; or greater than 1, indicating more important; among them, the green space density is quantified according to the normalized difference vegetation index, and its calculation formula is:

[0085]

[0086] Among them, NDVI is the normalized difference vegetation index, NIR is the reflection value in the near-infrared band, and R is the reflection value in the red light band.

[0087] In this embodiment, the finally constructed judgment matrix structure is as follows:

[0088] Factor Air humidity Precipitation Green space density Water level Air humidity 1 1 / 3 1 / 5 1 / 7 Precipitation 3 1 1 / 3 1 / 5 Green space density 5 3 1 1 / 3 Water level 7 5 3 1

[0089] In this matrix, for example, the importance of "air humidity" relative to "precipitation" is 1 / 3, meaning that "precipitation" is three times more important than "air humidity". Similarly, the importance of "water level" relative to "air humidity" is 7, meaning that "water level" is seven times more important than "air humidity".

[0090] In step S2, by calculating the eigenvalues and eigenvectors of the judgment matrix, the relative weight of each factor can be obtained. These weights reflect the importance of each factor in the risk assessment. The specific steps include:

[0091] First, find the eigenvector of the judgment matrix. The eigenvector is an attribute of the matrix and satisfies the following equation:

[0092] Ax = cx;

[0093] Where A is the judgment matrix, c is the eigenvalue, and x is the eigenvector;

[0094] After finding the eigenvector, normalize it so that the sum of its elements is 1. The normalized eigenvector is the relative weight of each factor.

[0095] In step S4 of this embodiment, according to the weight of each factor and the score of the evaluation object, calculate the comprehensive evaluation score, which reflects the overall performance of the evaluation object in the risk assessment. First, conduct data analysis based on the criterion of water level to determine the scores of different schemes. Among them, the score for the low water level is the lowest, and the score for the extremely high water level is the highest.

[0096] This embodiment uses a ten-point system to determine the scores of different schemes as follows: 2 points for the low water level, 4 points for the medium water level, 7 points for the high water level, and 9 points for the extremely high water level. Then, multiply the score of each scheme by the weight of the corresponding criterion to calculate the weighted score. The formula is as follows:

[0097] C = W × S;

[0098] where C represents the weighted score, W represents the weight, and S represents the scheme score.

[0099] In this embodiment, the calculation results of the weighted scores are as follows:

[0100] · Low water level scheme: 1.13;

[0101] · Medium water level scheme: 2.26;

[0102] · High water level scheme: 3.955;

[0103] · Extremely high water level scheme: 5.085.

[0104] Finally, add the flood control gate height index. Determine a benchmark value based on the analysis of historical flood data, and determine the adjustment of the flood control gate height under different water level schemes. The formula for calculating the flood control gate height is as follows:

[0105] H = C × Standard;

[0106] where H represents the flood control gate height determined under different water level schemes, C represents the weighted score, and Standard represents the flood control gate height benchmark, and calculate the flood control gate height determined under different water level schemes.

[0107] In this embodiment, the calculation results of the flood control gate height determined under different water level schemes are as follows:

[0108] · Flood control gate height for the low water level scheme: 339 mm;

[0109] · Flood control gate height for the medium water level scheme: 678 mm;

[0110] · Flood control gate height for the high water level scheme: 1186.5 mm;

[0111] · Flood control gate height for the extremely high water level scenario: 1525.5 mm,

[0112] In step S5 of this embodiment, according to the flood control gate heights determined under different water level scenarios calculated in step S3, in order to ensure the high efficiency of calculating the flood control gate heights determined under different water level scenarios and the production feasibility, and at the same time to ensure the accuracy of the data, the height data is rounded off in the tens place to correct the flood control gate height, and it is divided into four different levels of gate rising scenarios: low, medium, high, and extremely high. The four levels are specifically as follows:

[0113] (1) Low level:

[0114] Description: The rising height of the flood control gate is relatively low, applicable to low-risk flood events, and the corresponding weighted score is 1.13 - 2.26.

[0115] Feature: The rising height of the gate is 300 mm, which can cope with minor flood situations.

[0116] Applicable scenario: Applicable to areas with low flood risk, or as a preliminary flood control measure.

[0117] (2) Medium level:

[0118] Description: The rising height of the flood control gate is moderate, applicable to medium-risk flood events, and the corresponding weighted score is 2.26 - 3.955.

[0119] Feature: The rising height of the gate is 700 mm, which can cope with medium-intensity floods.

[0120] Applicable scenario: Applicable to areas with moderate flood risk, or as a conventional flood control measure.

[0121] (3) High level:

[0122] Description: The rising height of the flood control gate is relatively high, applicable to high-risk flood events, and the corresponding weighted score is 3.955 - 5.085.

[0123] Feature: The rising height of the gate is 1200 mm, which can cope with strong floods.

[0124] Applicable scenario: Applicable to areas with high flood risk, or as an emergency flood control measure.

[0125] (4) Extremely high level:

[0126] Description: The rising height of the flood control gate is extremely high, applicable to extremely high-risk flood events, and the corresponding weighted score is 5.085 - +∞.

[0127] Feature: The rising height of the gate is 1500 mm, which can cope with extreme flood situations.

[0128] Applicable scenarios: Applicable to areas with extremely high flood risks or as flood control measures in extreme situations.

[0129] As Figure 2 shown, this figure is the dynamic change state flow chart of the subway entrance and exit lifting gates. The cuboid represents the entrance and exit model of the subway, and the rectangle represents the gate sample diagrams of each module. Each stage is connected by an arrow, and the transfer control state is calculated and analyzed through four factors: air humidity, precipitation, green space density, and water level. The process from state one to state four represents the rising and falling trend of the gates.

[0130] Example Two

[0131] In this example, add a sample data set as follows:

[0132]

[0133] As Figure 3 and Figure 4 shown, in this example, the weights are calculated by the decision tree algorithm according to the information entropy and information gain theory in information theory, and the weight calculation is combined with the analytic hierarchy process. The specific weight calculation includes the following sub-steps:

[0134] Step S31: First, calculate the total information entropy, and the calculation formula is:

[0135]

[0136] where pi represents the sample proportion of category i, that is, the number of samples in category i divided by the total number of samples; n represents the number of label types; the calculation result will obtain the overall entropy value of the data before division;

[0137] Step S32: Calculate the information gain:

[0138] Traverse each feature f and calculate all possible values V(f) of the feature;

[0139] For each value v ∈ V(f), divide the data set S into sub-data sets Sv according to the value v of the feature f;

[0140] Calculate the entropy H(Sv) of each sub-data set;

[0141] Calculate the weighted average entropy after splitting, and the calculation formula is:

[0142]

[0143] Calculate the information gain, and the calculation formula is:

[0144] IG(S,f) = H(S) - weighted average entropy;

[0145] Step S33: Select the optimal splitting feature:

[0146] Compare the information gains of all features and select the feature with the largest gain as the splitting feature of the current node;

[0147] Step S34: Recursive splitting:

[0148] Recursively call the above steps for each value of the selected splitting feature to continue constructing child nodes until all features are used up;

[0149] Step S35: Output the importance of features. The feature importance measures the contribution degree of each feature in the model, mainly based on the impurity reduction when splitting nodes;

[0150]

[0151] Step S36: Output the result after importance normalization.

[0152] In this embodiment, the weighted score formula is changed to

[0153] where C represents the weighted score, W represents the analytic hierarchy process weight, I represents the importance calculated by the decision tree, I is the result of importance normalization output in the above step S36, and S represents the scheme score; this formula combines the feature importance output by the decision tree model and the weight calculated based on the judgment matrix for comprehensive weight adjustment, which can optimize the hierarchical structure and improve the comprehensive decision-making ability.

[0154] The calculation of the weighted score result is as follows:

[0155] · Low water level scheme: 1.05;

[0156] · Medium water level scheme: 2.13;

[0157] · High water level scheme: 4.008;

[0158] · Extremely high water level scheme: 5.029.

[0159] The calculation of the remaining height is exactly the same as that in Embodiment 1. The difference between Embodiment 2 and Embodiment 1 is that Embodiment 1 directly uses the analytic hierarchy process to calculate the comprehensive score. While Embodiment 2 adds a decision tree model on the basis of the analytic hierarchy process, combines the result of importance normalization calculated by the decision tree model and the weight of the analytic hierarchy process to calculate the weighted score. The decision tree algorithm in Embodiment 2 is more inclined to be used when there is a reasonable amount of data in the dataset. If there is no dataset or the amount of the dataset is insufficient, then the method of Embodiment 1 can be directly adopted for calculation.

[0160] Embodiment 3

[0161] Based on the first and second embodiments, this embodiment further adds a manual control mode to the overall solution. The manual control mode is manually controlled by a human, and the rising and falling heights of the gates at each stage can be selected according to the actual situation, thereby increasing the flexibility of implementation.

[0162] It should be understood that each part of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits with logic gate circuits for implementing logical functions on data signals, application specific integrated circuits with appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0163] Those of ordinary skill in the art of this technology can understand that all or part of the steps carried by the methods in the above embodiments can be completed by instructing relevant hardware through a program. The said program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0164] In addition, in each embodiment of the present invention, the functional units can be integrated into a processing module, or each unit can exist physically alone, or two or more units can be integrated into one module. The above integrated module can be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0165] The above-mentioned storage medium can be a read-only memory, a magnetic disk, an optical disk, etc.

[0166] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the purpose and scope of the present technical solution, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for controlling a flood control reversible gate, characterized in that: The method includes automatic control of the gate rise, that is, by collecting air humidity, precipitation, green space density, and water level data as factors and jointly analyzing their respective weights, obtaining water level weights and dividing interval weights, setting gate rise plans of different levels of flood control gate lifting heights after comprehensive analysis, and determining the gate rise plan and transmitting the gate flip rise signal based on the water level rise data of the water level trigger collected in real time, thereby controlling the gate rise.

2. A method for controlling a flood control reversible gate according to claim 1, characterized in that: The method also includes automatic control of gate descent, that is, controlling the descent of the gate through water level drop data collected in real time by a water level trigger.

3. A method for controlling a flood control reversible gate according to claim 1, characterized in that: The automatic control of the gate's rise specifically includes the following steps: Step S1: Decompose the flood gate height adjustment problem into multiple levels, including the target level, criterion level, sub-criterion level and scheme level, and construct a judgment matrix: Step S2: By calculating the eigenvalues ​​and eigenvectors of the judgment matrix, the relative weight of each factor can be obtained; Step S3: Calculate the comprehensive evaluation score according to the relative weight of each factor and the score of the evaluation object, and calculate the height of the flood control gate determined under different water level schemes; Step S4: Divide the gate raising plans into different levels according to the comprehensive evaluation scores.

4. A method for controlling a flood control reversible gate according to claim 3, characterized in that: In step S3, before calculating the comprehensive evaluation score, the importance of each factor is obtained and normalized by calculating the information entropy and information gain of the data set; It includes the following sub-steps: Step S31: First, calculate the total information entropy, the calculation formula is: Among them, pi represents the sample ratio of category i, that is, the number of samples of category i divided by the total number of samples; n represents the number of label types; the calculation result will get the overall entropy value of the data before division; Step S32: Calculate information gain: Traverse each feature f and calculate all possible values ​​V(f) of the feature; For each value v∈V(f), divide the data set S into sub-data sets Sv according to the value v of feature f; Calculate the entropy H(Sv) of each sub-dataset; Calculate the weighted average entropy after splitting, the calculation formula is: Calculate the information gain, the calculation formula is: IG(S,f)=H(S)-weighted average entropy; Step S33: Select the optimal splitting feature: Compare the information gain of all features and select the feature with the largest gain as the split feature of the current node; Step S34: Recursive splitting: Recursively call the above steps for each value of the selected split feature, and continue to build child nodes until all features are used up; Step S35: output the importance of the feature. The feature importance measures the contribution of each feature in the model, mainly based on the reduction of impurity when splitting the node; Step S36: Output the importance normalized result.

5. A method for controlling a flood control reversible gate according to claim 3, characterized in that: In step S1, the target layer is divided into the height of the flood gate for dynamic adjustment; the criterion layer is divided into air humidity, precipitation, green space density, and water level; the sub-criterion layer is divided into air humidity and precipitation, and the air humidity includes relative humidity, absolute humidity, and humidity change rate; the precipitation includes rainfall, precipitation intensity, and precipitation frequency; the scheme layer is divided into different schemes for the lifting height of the flood gate, including four levels: low, medium, high, and extremely high.

6. A method for controlling a flood control flip gate according to claim 3, characterized in that: In step S1, the judgment matrix is ​​constructed by comparing the elements of each level by two-by-two comparisons. The elements in the judgment matrix represent the relative importance of two factors. The weight of each factor relative to the previous factor is 1, indicating equal importance; or greater than 1, indicating more important. The green space density is quantified according to the normalized vegetation index, and its calculation formula is: Among them, NDVI is the normalized vegetation index, NIR is the reflectance value of the near-infrared band, and R is the reflectance value of the red light band.

7. A method for controlling a flood control reversible gate according to claim 3, characterized in that: In step S2, the importance of each factor in risk assessment is reflected by weight, and the specific steps include: Find the eigenvectors of the judgment matrix. An eigenvector is a property of a matrix that satisfies the following equation: Ax = cx; Where A is the judgment matrix, c is the eigenvalue, and x is the eigenvector; After finding the eigenvector, normalize it so that the sum of its elements is 1. The normalized eigenvector is the relative weight of each factor.

8. A method for controlling a flood control tilt gate according to claim 3 or 4, characterized in that: In step S3, a comprehensive evaluation score is calculated based on the weight of each factor and the score of the evaluation object, and the comprehensive evaluation score reflects the overall performance of the evaluation object in the risk assessment; First, data analysis was performed based on the water level criterion to determine the scores of different schemes, with low water levels having the lowest scores and very high water levels having the highest scores; Then, multiply the score of each solution by the weight of the corresponding criterion, and calculate the weighted score formula as follows: Where C represents the weighted score; W represents the weight; I represents the importance calculated by the decision tree; S represents the solution score; Finally, the flood gate height index is added, and a benchmark value is determined based on the historical flood data analysis to determine the final flood gate height adjustment under different water level schemes. The formula for calculating the flood gate height is as follows: H = C × Standard; Where H represents the height of the flood gate determined under different water level scenarios, C represents the weighted score, and Standard represents the flood gate height benchmark. The height of the flood gate determined under different water level scenarios is calculated.

9. A method for controlling a flood control reversible gate according to claim 3, characterized in that: In step S3, according to the calculated flood gate heights determined under different water level schemes, the flood gate heights are corrected and divided into four different levels of gate raising schemes: low, medium, high, and extremely high.

10. A method for controlling a flood control reversible gate according to claim 1, characterized in that: The method also includes a manual control mode of the gate, wherein the manual control mode is manually controlled by humans, and the rising and falling heights of the gate in each stage are selected by humans.

Images