Safety evaluation method for engineering safety monitoring system

Abstract

The invention relates to the technical field of engineering safety evaluation, and discloses an engineering safety monitoring system safety evaluation method, which comprises the steps of data acquisition, preprocessing, model building and health evaluation. According to the invention, multiple types of sensors are arranged at key positions of the hydraulic structure according to a principle, acquisition frequency is set according to characteristics, data can be comprehensively and accurately acquired, and redundancy is avoided; abnormal values are judged and replaced by a scientific method and a reasonable preset multiple, and the data quality is improved; when the model is built, a regression equation is built by utilizing historical monitoring data through rigorous mathematical statistics steps, and the relation of all variables is accurately reflected, so that the model is scientific and accurate; meanwhile, an actual osmotic pressure correlation equation is constructed, the health state is judged by comparing the error of a predicted value and an actual value and referring to a threshold value, and hidden danger can be early warned in time. On the whole, all links cooperate to form a complete evaluation process, reliable guarantee is provided for safe operation of the hydraulic construction, and risks are effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of engineering safety evaluation, and particularly relates to a safety evaluation method for an engineering safety monitoring system. Background Art

[0002] In the field of water conservancy projects, the safe operation of hydraulic structures is of crucial importance. With the continuous expansion of the scale of water conservancy projects and the increase in the operation time, the safety risks they face are becoming increasingly complex and diverse. Traditional safety monitoring methods for hydraulic structures often have many limitations.

[0003] In the face of a large amount of monitoring data, due to the lack of a scientific and reasonable outlier removal mechanism, abnormal data may be mixed into the normal data sequence, interfering with the judgment of the true state of the structure. At the same time, the existing model construction methods are not rigorous and perfect enough, and do not fully consider the complex internal relationship between environmental quantities and monitoring effect quantities. The constructed models are difficult to accurately simulate the actual operation laws of hydraulic structures under different working conditions, greatly reducing the accuracy of safety assessment results; at the same time, they cannot timely and accurately judge the health status of hydraulic structures based on monitoring data and give reliable early warning information. This makes it difficult to take effective preventive measures before a safety accident occurs. Once an accident occurs, it often causes huge economic losses, casualties, and serious damage to the surrounding environment.

[0004] In summary, there is an urgent need for a more scientific, comprehensive, and accurate safety evaluation method for engineering safety monitoring systems to improve the reliability and effectiveness of safety monitoring and assessment of hydraulic structures, ensure the safe and stable operation of water conservancy projects, and reduce safety risks and accident losses. Summary of the Invention

[0005] The purpose of the present invention is to provide a safety evaluation method for an engineering safety monitoring system, which solves the technical problems proposed in the background art.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A safety evaluation method for an engineering safety monitoring system includes the following steps:

[0008] The first step, data acquisition

[0009] Use a variety of sensors to obtain the target monitoring data of the hydraulic structure;

[0010] The second step, preprocessing

[0011] Perform outlier removal processing on the target monitoring data collected from the hydraulic structure;

[0012] The third step, model construction

[0013] Construct a monitoring and statistical model based on the historical monitoring data extracted in the historical monitoring period;

[0014] Step 4, Health assessment

[0015] Substitute the target monitoring data into the constructed monitoring and statistical model, determine the predicted values of the corresponding data, and then determine the health status of the hydraulic structure within the specified period based on the difference between the predicted values and the actual monitoring values of the corresponding data.

[0016] As a further solution of the present invention: In the data acquisition step, a pressure sensor, a water level sensor, and a temperature sensor are specifically used and arranged at key positions of the hydraulic structure, and then the parameters corresponding to the engineering seepage pressure, water level, and temperature are monitored in real time, and a seepage pressure data sequence, a water level data sequence, and a temperature data sequence are obtained.

[0017] As a further solution of the present invention: The outlier removal processing method is as follows:

[0018] Within the specified period, obtain the seepage pressure data sequence, the water level data sequence, and the temperature data sequence, and mark them as P i , H i and T i , where i = 1, 2,..., n, and n represents the number of acquisition points within the specified period;

[0019] Then calculate the average values corresponding to P i , H i and T i respectively, and record them as PP, HP, and TP in sequence;

[0020] At the same time, calculate the standard deviations corresponding to P i , H i and T i respectively, and record them as PC, HC, and TC in sequence;

[0021] Then compare |P i -PP| with q times of PC, compare |H i -HP| with q times of HC, and compare |T i -TP| with q times of TC:

[0022] If |P i -PP| > q×PC, it means that the corresponding P i is an outlier. Then obtain P i -1 and P i +1 adjacent to it before and after, and then calculate the average value between P i -1 and P i +1, and use it as the outlier P i +1 iReplacement value;

[0023] If |H i - HP| > q × HC, it means that the corresponding Hi is an outlier, and then obtain the H i adjacent to it before and after i -1 and H i +1, and then calculate the average value between H i -1 and H i +1, and use it as the replacement value of the outlier Hi;

[0024] If |T i - TP| > q × TC, it means that the corresponding T i is an outlier, and then obtain the T i adjacent to it before and after i -1 and T i +1, and then calculate the average value between T i -1 and T i +1, and use it as the replacement value of the outlier T i ;

[0025] If |P i - PP| ≤ q × PC, it is determined that the corresponding P i value is normal;

[0026] If |H i - HP| ≤ q × HC, it is determined that the corresponding H i value is normal;

[0027] If |Ti - TP| ≤ q × TC, it is determined that the corresponding T i value is normal;

[0028] Among them, q is a preset multiple value;

[0029] When the i value corresponding to the outlier in the corresponding data sequence is 1 or n, the average value in the corresponding data sequence is used as the replacement value of the outlier in the corresponding data sequence.

[0030] As a further solution of the present invention: The construction method of the monitoring statistical model is as follows:

[0031] StepA1. In the historical monitoring period, extract the corresponding historical monitoring data;

[0032] The historical monitoring data includes:

[0033] Environmental quantities, including but not limited to temperature, water level;

[0034] Monitoring effect quantities, including but not limited to seepage pressure;

[0035] Step A2. Subsequently, using the environmental quantity as the independent variable and the monitoring effect quantity as the dependent variable, establish a regression equation between the two using mathematical statistics analysis methods;

[0036] Among them, the regression equation is the core component of the monitoring statistical model, which expresses the quantitative relationship between the independent variable and the dependent variable in the form of a mathematical formula.

[0037] As a further solution of the present invention: the regression equation is established as follows:

[0038] Step A2.1. Obtain multiple sample values corresponding to each independent variable and label them as X tk , where t = 1, 2,..., e, k = 1, 2,..., r, e represents the number of independent variables, and r represents the number of multiple sample values in each independent variable;

[0039] At the same time, obtain multiple sample values corresponding to each dependent variable and label them as Y hk , where h = 1, 2,..., w, k = 1, 2,..., r, h represents the number of dependent variables, and r represents the number of multiple sample values in each dependent variable;

[0040] Step A2.2. Select one independent variable from multiple independent variables and one dependent variable from multiple dependent variables, and calculate the correlation coefficient between the independent variable and the dependent variable respectively;

[0041] The calculation formula of the correlation coefficient is as follows:

[0042] Calculate the correlation coefficient R k between the independent variable Xa k and the dependent variable Yb ca ;

[0043] In the formula, Xa k ∈X tk , Yb k ∈Y hk ;

[0044] XPa is the average value of all Xa k , and YPc is the average value of all Yc k ;

[0045] And so on, calculate the correlation coefficients between all independent variables and dependent variables, and obtain the correlation coefficient matrix;

[0046] Step A2.3. At the same time, among multiple independent variables, calculate the correlation coefficient between two independent variables respectively;

[0047] The calculation formula of the correlation coefficient is as follows:

[0048] Calculate the independent variable Xa1 k and Xa2 k The correlation coefficient R a1a2 ;

[0049] Wherein, Xa1 k ∈X tk , Xa2 k ∈X tk , and Xa1 k ≠Xa2 k ;

[0050] XPa1 is the average value of all Xa1 k , XPa2 is the average value of all Xa2 k ;

[0051] StepA2.4. Find the independent variable corresponding to the correlation coefficient with the largest absolute value from the correlation coefficient matrix obtained in StepA2.2, and use it as the first factor introduced into the regression equation;

[0052] And after introducing the first factor, calculate the partial correlation coefficient between the remaining independent variables and the dependent variable in the presence of the introduced factor;

[0053] The calculation method of the partial correlation coefficient is as follows:

[0054]

[0055] Wherein, R c(a1a2) is the partial correlation coefficient, R ca1 ∈R ca , R ca2 ∈R ca , and R ca1 ≠R ca2 ;

[0056] StepA2.5. Further determine whether to select the factor with the highest degree of correlation with the dependent variable from the remaining factors and introduce it into the regression equation according to the magnitude of the partial correlation coefficient;

[0057] If the partial correlation coefficient is significant, introduce it into the equation;

[0058] If it is not significant, do not introduce it temporarily and continue to consider the relationship between other independent variables and the dependent variable.

[0059] As a further solution of the present invention: Among them, whether the partial correlation coefficient is significant is determined by a significance test.

[0060] As a further solution of the present invention: The significance test method is as follows:

[0061] It is based on each introduced factor;

[0062] StepP1, through:

[0063] Calculate the regression sum of squares SSR;

[0064] Among them, Y0 h is the mean of all sample values ​​of the dependent variable, Yy hk is the predicted value calculated from the regression equation, and the method used to obtain this predicted value includes the introduced independent variables;

[0065] Step P2, through:

[0066] Calculate the regression sum of squares SSR;

[0067] Among them, Y hk is the actual observed value;

[0068] Step P3, through: SST = SSR + SSE

[0069] Calculate the total sum of squares;

[0070] Step P4, through:

[0071] Calculate the test statistic F;

[0072] Among them, u is the number of independent variables introduced;

[0073] StepP5, extract the preset significance level α;

[0074] Then compare F with Fα(k,mk-1):

[0075] If F>Fα(k,mk-1), the factor is significant and is retained in the regression equation;

[0076] If F≤Fα(k, mk-1), the factor is not significant and is eliminated, and the regression equation is recalculated until all introduced factors pass the significance test.

[0077] As a further solution of the present invention: wherein α is a probability value pre-set in the hypothesis test, which is used to determine whether the evidence provided by the sample data is sufficient to reject the null hypothesis, and Fα(k, mk-1) is the critical value of the F distribution, which is a specified value calculated according to a distribution table or by statistical software.

[0078] As a further solution of the present invention: the health assessment method is as follows:

[0079] Step G1, according to the establishment method of regression equation, construct the regression equation related to seepage pressure, water level and temperature;

[0080] The expression form of its regression equation is as follows:

[0081] P = b 0 + b 1 × H + b 2 × T + b 3 × H 2 + b 4 × T 2 + b 5 × H × T;

[0082] where P is the seepage pressure, H is the water level, T is the temperature, and b 0 , b 1 , b 2 , b 3 , b 4 , b 5 are regression coefficients, and the regression coefficients are determined values obtained by performing regression analysis on historical monitoring data;

[0083] StepG2. Substitute the average values corresponding to the real-time monitored water level data sequence and temperature data sequence within the specified period into the regression equation, and calculate the seepage pressure prediction value;

[0084] The calculation method is as follows:

[0085] Py = b 0 + b 1 × HP + b 2 × TP + b 3 × HP 2 + b 4 × TP 2 + b 5 × HP × TP;

[0086] where Py is the seepage pressure prediction value, H is the average value of the water level data sequence, and T is the average value of the temperature data sequence;

[0087] StepG3. Take the average value of the seepage pressure data sequence within the specified period as the actual monitored seepage pressure value;

[0088] StepG4. Calculate the error wc between the seepage pressure prediction value and the actual monitored seepage pressure value through: wc = PP - Py;

[0089] Then compare the absolute value of its error with the corresponding preset error threshold wcy:

[0090] If |wc| ≤ wcy, it is determined that the health state of the hydraulic structure within the specified period is normal;

[0091] If |wc| > wcy, it is determined that the health status of the hydraulic structure is abnormal within the specified period, that is, there are potential safety hazards in the hydraulic structure within the specified period, and a warning signal is generated accordingly.

[0092] Advantages of the present invention:

[0093] In the present invention, by adopting a variety of sensors such as pressure sensors, water level sensors, and temperature sensors and arranging them at key positions of the hydraulic structure, it is possible to monitor in real time the parameters corresponding to the seepage pressure, water level, and temperature of the project, and obtain multi-dimensional target monitoring data such as seepage pressure data sequences, water level data sequences, and temperature data sequences, laying a solid data foundation for comprehensively and accurately evaluating the health status of the hydraulic structure subsequently.

[0094] The present invention has a detailed and scientific and reasonable method for removing outliers. Within the specified period, the average value and standard deviation are calculated respectively for different data sequences, and through a comparison rule related to a preset multiple value, it is accurately determined whether the data is an outlier, and a corresponding appropriate replacement value scheme is given for outliers at different positions, making the data entering the subsequent analysis more reliable and reducing the interference of abnormal data on the overall evaluation result.

[0095] In the present invention, when constructing the monitoring statistical model, the environmental quantity and monitoring effect quantity in the historical monitoring period are fully considered. Taking the environmental quantity as the independent variable and the monitoring effect quantity as the dependent variable, a regression equation is established by using mathematical statistics analysis methods, so that the quantitative relationship between the independent variable and the dependent variable can be accurately expressed in the form of a mathematical formula, and more scientifically reflect the influence of the association between various factors on the hydraulic structure.

[0096] In the present invention, the establishment of the regression equation has rigorous steps. From calculating the correlation coefficients between each independent variable and the dependent variable, and the correlation coefficients between independent variables, to finding the key introduced factors based on the correlation coefficients, and further screening the introduced factors through partial correlation coefficients, and combining significance tests to ensure that the factors introduced into the regression equation are all significant and reasonable, making the finally constructed monitoring statistical model have high accuracy and scientificity and can effectively simulate the actual situation.

[0097] In the present invention, a regression equation related to seepage pressure, water level, and temperature is constructed according to the actual situation, and the regression coefficients are obtained as determined values through regression analysis of historical monitoring data, so that the evaluation model can fit the actual operation status of the hydraulic structure to predict the seepage pressure situation.

[0098] In the present invention, the average value of real-time monitoring data is substituted into the regression equation to obtain the seepage pressure prediction value, which is then compared with the average value of the actual monitored seepage pressure value. Based on the comparison result between the error and the preset error threshold, it can accurately determine whether the health state of the hydraulic structure is normal or abnormal within the specified period. Once an abnormality is detected, it is determined that there is a potential safety hazard and an early warning signal can be generated in a timely manner, which helps to take corresponding measures in advance to ensure the safety of the hydraulic structure and effectively avoid the occurrence of safety accidents.

[0099] Generally speaking, the safety evaluation method of the engineering safety monitoring system in this project is closely coordinated and scientifically reasonable in all aspects from data collection, preprocessing, model construction to health assessment. It can relatively accurately evaluate and warn the safety state of hydraulic structures, providing a strong guarantee for the safe operation of hydraulic structures. Brief Description of the Drawings

[0100] The present invention will be further described below with reference to the accompanying drawings.

[0101] Figure 1 It is a schematic flow chart of a safety evaluation method for an engineering safety monitoring system of the present invention.

[0102] Figure 2 It is a schematic flow chart of the calculation process of the regression equation in a safety evaluation method for an engineering safety monitoring system of the present invention. Detailed Embodiments

[0103] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts shall fall within the protection scope of the present invention.

[0104] Embodiment 1

[0105] Please refer to Figure 1 and Figure 2 As shown, the present invention is a safety evaluation method for an engineering safety monitoring system, including the following steps:

[0106] The first step: Data collection

[0107] A variety of sensors are used to obtain the target monitoring data of the hydraulic structure;

[0108] Specifically, pressure sensors, water level sensors, and temperature sensors are used and arranged at key positions of the hydraulic structure, and then the parameters corresponding to the engineering seepage pressure, water level, and temperature are monitored in real time, and a seepage pressure data sequence, a water level data sequence, and a temperature data sequence are obtained;

[0109] In this embodiment, the arrangement of sensors follows the principles of comprehensiveness and representativeness. Sensors are reasonably arranged at key positions of hydraulic structures, such as the dam body, dam foundation, spillway, etc. For large dams, a row of sensors is arranged at regular intervals along the dam axis direction, such as every 50 m, and monitoring sections are set at different elevations of the dam body, such as every 10 m, to ensure that various parameter changes of the engineering structure can be comprehensively monitored.

[0110] In this embodiment, the data acquisition frequencies for engineering seepage pressure, water level, and temperature are the same. The data acquisition frequency is also set according to the importance of the project and the structural characteristics. For key positions or parameters that change rapidly, such as dam seepage monitoring, the acquisition frequency can be set to once every 10 minutes; for relatively stable parameters, such as water temperature monitoring, the acquisition frequency can be once every hour, in order to obtain sufficient valid data while avoiding data redundancy.

[0111] Step 2: Pretreatment

[0112] Perform outlier removal processing on the target monitoring data collected from hydraulic structures.

[0113] The specific method is as follows:

[0114] Within a specified period, obtain the seepage pressure data sequence, water level data sequence, and temperature data sequence, and mark them as P i , H i and T i , where i = 1, 2,..., n, and n represents the number of acquisition points within the specified period.

[0115] Subsequently, calculate the average values corresponding to P i , H i and T i respectively, and record them as PP, HP, and TP in sequence.

[0116] At the same time, calculate the standard deviations corresponding to P i , H i and T i respectively, and record them as PC, HC, and TC in sequence.

[0117] Then compare |P i - PP| with q times PC, compare |H i - HP| with q times HC, and compare |T i - TP| with q times TC:

[0118] If |P i - PP| > q × PC, it means that the corresponding P i is an outlier. Subsequently, obtain the P i adjacent to and before and after this P i -1 and Pi Increment by 1, and then calculate P i Decrement by 1 and P i Calculate the average value between increment by 1 and use it as the replacement value for the outlier P i ;

[0119] If |H i - HP| > q × HC, it indicates that the corresponding Hi is an outlier. Then obtain the H i before and after this H i decrement by 1 and H i increment by 1, and then calculate the H i decrement by 1 and H i increment by 1. Calculate the average value between them and use it as the replacement value for the outlier Hi;

[0120] If |T i - TP| > q × TC, it indicates that the corresponding T i is an outlier. Then obtain the T i before and after this T i decrement by 1 and T i increment by 1, and then calculate the T i decrement by 1 and T i increment by 1. Calculate the average value between them and use it as the replacement value for the outlier T i ;

[0121] If |P i - PP| ≤ q × PC, it is determined that the corresponding P i value is normal;

[0122] If |H i - HP| ≤ q × HC, it is determined that the corresponding H i value is normal;

[0123] If |Ti - TP| ≤ q × TC, it is determined that the corresponding T i value is normal;

[0124] Among them, q is a preset multiple value. In this embodiment, q is taken as 3;

[0125] When the i value corresponding to the outlier in the corresponding data sequence is 1 or n, then use the average value in the corresponding data sequence as the replacement value for the outlier in the corresponding data sequence;

[0126] In this embodiment, by arranging pressure sensors, water level sensors, and temperature sensors at key positions of hydraulic structures according to the principles of comprehensiveness and representativeness, parameter data sequences corresponding to engineering seepage pressure, water level, and temperature can be comprehensively and accurately obtained; a scientific outlier removal method is adopted, based on the average value and standard deviation of the data sequence, and combined with a preset multiple to judge outliers; outliers can be effectively identified and replaced, making the data entering subsequent analysis more reliable, reducing the interference of abnormal data on the overall evaluation result, improving data quality, and providing more accurate data support for the preliminary judgment of the state of hydraulic structures.

[0127] Embodiment 2

[0128] As Embodiment 2 of the present invention, in the specific implementation of this application, compared with Embodiment 1, the technical solution of this embodiment is only different from that of Embodiment 1 in that this embodiment further includes the step: model construction

[0129] Construct a monitoring statistical model based on the historical monitoring data extracted in the historical monitoring period;

[0130] The method is as follows:

[0131] StepA1. Extract the corresponding historical monitoring data in the historical monitoring period;

[0132] The historical monitoring data includes:

[0133] Environmental quantities, including but not limited to temperature and water level;

[0134] Monitoring effect quantities, including but not limited to seepage pressure;

[0135] StepA2. Then, use the environmental quantity as the independent variable, the monitoring effect quantity as the dependent variable, and apply the mathematical statistics analysis method to establish a regression equation between the two;

[0136] The establishment method of the regression equation is as follows:

[0137] StepA2.1. Obtain multiple sample values corresponding to each independent variable and mark them as X tk , where t = 1, 2,..., e, k = 1, 2,..., r, e represents the number of independent variables, and r represents the number of multiple sample values in each independent variable;

[0138] At the same time, obtain multiple sample values corresponding to each dependent variable and mark them as Y hk , where h = 1, 2,..., w, k = 1, 2,..., r, h represents the number of dependent variables, and r represents the number of multiple sample values in each dependent variable;

[0139] StepA2.2. Select one independent variable from multiple independent variables and one dependent variable from multiple dependent variables, and calculate the correlation coefficient between the independent variable and the dependent variable respectively;

[0140] The calculation formula of the correlation coefficient is as follows:

[0141] Calculate the correlation coefficient R k between the independent variable Xa k and the dependent variable Yb ca ;

[0142] In the formula, Xa k ∈X tk , Yb k ∈Y hk ;

[0143] XPa is the average value of all Xa k , and YPc is the average value of all Yc k ;

[0144] And so on, calculate the correlation coefficients between all independent variables and dependent variables, and obtain the correlation coefficient matrix;

[0145] StepA2.3. At the same time, among multiple independent variables, calculate the correlation coefficients between two independent variables respectively;

[0146] The calculation formula of the correlation coefficient is as follows:

[0147] Calculate the correlation coefficient R k between the independent variables Xa1 k and Xa2 a1a2 ;

[0148] In the formula, Xa1 k ∈X tk , Xa2 k ∈X tk , and Xa1 k ≠Xa2 k ;

[0149] XPa1 is the average value of all Xa1 k , and XPa2 is the average value of all Xa2 k ;

[0150] StepA2.4. Find the independent variable corresponding to the correlation coefficient with the largest absolute value from the correlation coefficient matrix obtained in StepA2.2, and use it as the first factor introduced into the regression equation;

[0151] After introducing the first factor, calculate the partial correlation coefficients between the remaining independent variables and the dependent variable in the presence of the introduced factor;

[0152] The calculation method of the partial correlation coefficient is as follows:

[0153]

[0154] In the formula, R c(a1a2) is the partial correlation coefficient, R ca1 ∈R ca and R ca2 ∈R ca and R ca1 ≠R ca2 ;

[0155] StepA2.5. Further determine whether to introduce the factor with the highest degree of correlation with the dependent variable from the remaining factors into the regression equation according to the magnitude of the partial correlation coefficient;

[0156] If the partial correlation coefficient is significant, introduce it into the equation;

[0157] If it is not significant, do not introduce it temporarily and continue to consider the relationship between other independent variables and the dependent variable;

[0158] Among them, whether the partial correlation coefficient is significant is determined by a significance test:

[0159] The significance test method is as follows:

[0160] It is based on each introduced factor;

[0161] StepP1. Through:

[0162] Calculate the regression sum of squares SSR;

[0163] Among them, Y0 h is the mean of all sample values corresponding to the dependent variable, Yy hk is the predicted value calculated according to the regression equation, and the method of obtaining this predicted value includes the introduced independent variables;

[0164] StepP2. Through:

[0165] Calculate the regression sum of squares SSR;

[0166] Among them, Y hk is the actual observed value;

[0167] StepP3. Through: SST = SSR + SSE

[0168] Calculate the total sum of squares;

[0169] StepP4. Through:

[0170] Calculate the test statistic F;

[0171] where u is the number of independent variables already introduced;

[0172] StepP5: Extract the preset significance level α;

[0173] Then compare F with Fα(k, m - k - 1):

[0174] If F > Fα(k, m - k - 1), then the factor is significant and is retained in the regression equation;

[0175] If F ≤ Fα(k, m - k - 1), then the factor is not significant and the factor is removed, and the regression equation is recalculated until all introduced factors pass the significance test;

[0176] where α is a probability value preset in the hypothesis test for judging whether the evidence provided by the sample data is sufficient to reject the null hypothesis, and Fα(k, m - k - 1) is the critical value of the F - distribution, which is a specified value obtained according to the distribution table or calculated by statistical software;

[0177] This embodiment adds a model - building step on the basis of Embodiment 1. By extracting the environmental quantity and monitoring effect quantity data in the historical monitoring period, it provides a comprehensive data source for constructing the monitoring statistical model. These historical data reflect the operation characteristics of the hydraulic structure under different working conditions, which helps to establish a more practical model; using the mathematical statistics analysis method to establish the regression equation between the environmental quantity and the monitoring effect quantity. During the establishment process, the correlation coefficient between the independent variable and the dependent variable and the correlation coefficient between the independent variables are calculated in detail. According to the correlation coefficient matrix, the factors introduced into the regression equation are determined, and the factors are further screened through the partial correlation coefficient and the significance test. Such a rigorous step ensures that the regression equation can accurately express the quantitative relationship between the independent variable and the dependent variable, making the constructed monitoring statistical model have high accuracy and scientificity, and can effectively simulate the complex internal relationship between the environmental quantity and the monitoring effect quantity of the hydraulic structure under different working conditions, laying a foundation for more accurate safety assessment.

[0178] Embodiment 3

[0179] As Embodiment 3 of the present invention, when this application is specifically implemented, compared with Embodiment 1 and Embodiment 2, the technical solution of this embodiment is to combine and implement the solutions of the above - mentioned Embodiment 1 and Embodiment 2. The difference between the technical solution of this embodiment and Embodiment 1 and Embodiment 2 is only that this embodiment further includes the step: Health assessment

[0180] StepG1: According to the establishment method of the regression equation, construct a regression equation related to seepage pressure, water level and temperature;

[0181] The expression form of its regression equation is as follows:

[0182] P = b 0 + b 1 × H + b 2 × T + b 3 × H 2 + b 4 × T 2 + b 5 × H × T;

[0183] Wherein, P is the seepage pressure, H is the water level, T is the temperature, and b 0 , b 1 , b 2 , b 3 , b 4 , b 5 are regression coefficients, and the regression coefficients are determined values obtained by performing regression analysis on historical monitoring data;

[0184] StepG2. Substitute the average values corresponding to the real-time monitored water level data sequence and temperature data sequence within the specified period into the regression equation, and calculate the seepage pressure prediction value;

[0185] The calculation method is as follows:

[0186] Py = b 0 + b 1 × HP + b 2 × TP + b 3 × HP 2 + b 4 × TP 2 + b 5 × HP × TP;

[0187] Wherein, Py is the seepage pressure prediction value, H is the average value of the water level data sequence, and T is the average value of the temperature data sequence;

[0188] StepG3. Take the average value of the seepage pressure data sequence within the specified period as the actual monitored seepage pressure value;

[0189] StepG4. Calculate the error wc between the seepage pressure prediction value and the actual monitored seepage pressure value through: wc = PP - Py;

[0190] Then compare the absolute value of its error with the corresponding preset error threshold wcy:

[0191] If |wc| ≤ wcy, it is determined that the health state of the hydraulic structure within the specified period is normal;

[0192] If |wc| > wcy, it is determined that the health status of the hydraulic structure is abnormal within the specified period, that is, there are potential safety hazards in the hydraulic structure within the specified period, and a warning signal is generated accordingly.

[0193] This embodiment adds a health assessment step on the basis of Embodiment 1 and Embodiment 2. A specific regression equation related to seepage pressure, water level, and temperature is constructed. This equation is based on the regression analysis of historical monitoring data to obtain determined regression coefficients, which can better reflect the actual situation. By substituting the average values of the real-time monitored water level and temperature data sequences into the regression equation to calculate the predicted seepage pressure value, and comparing the error with the average value of the actual monitored seepage pressure value, the health status of the hydraulic structure is determined according to the comparison result of the error and the preset error threshold. It can timely and accurately judge whether there are potential safety hazards in the hydraulic structure based on the monitoring data. Once an abnormality is found, a warning signal can be generated immediately, which helps to take corresponding measures in advance to ensure the safety of the hydraulic structure and effectively avoid the occurrence of safety accidents, improving the evaluation ability of the entire engineering safety monitoring system for the health status of the hydraulic structure and the timeliness of early warning.

[0194] Embodiment 4

[0195] As Embodiment 4 of the present invention, when this application is specifically implemented, compared with Embodiment 1, Embodiment 2, and Embodiment 3, the technical solution of this embodiment is to combine and implement the solutions of the above Embodiment 1, Embodiment 2, and Embodiment 3.

[0196] This embodiment integrates the solutions of Embodiment 1, 2, and 3, forming a complete safety evaluation process for the engineering safety monitoring system from data collection, preprocessing, model building to health assessment. Each link closely cooperates and connects with each other, giving full play to the advantages of each link, and realizing the comprehensive, accurate, and dynamic monitoring and evaluation of the safety status of the hydraulic structure. It can continuously provide accurate safety information during the operation of the hydraulic structure, provide all-round guarantee for the long-term stable operation of the project, effectively reduce safety risks, and improve the safety, reliability, and management efficiency of the water conservancy project.

[0197] The above formulas are all dimensionless and take their numerical values for calculation. The formula is obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters and threshold selection in the formula are set by those skilled in the art according to the actual situation.

[0198] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A safety evaluation method for an engineering safety monitoring system, characterized in that: The following steps are involved: Step 1: Data Collection Use a variety of sensors to obtain target monitoring data of hydraulic structures; Step 2: Preprocessing Perform outlier removal processing on the target monitoring data collected from hydraulic structures; Step 3: Model Building Construct a monitoring statistical model based on historical monitoring data extracted during the historical monitoring cycle; Step 4: Health Assessment The target monitoring data is substituted into the constructed monitoring statistical model, and the predicted value of the corresponding data is determined. Then, the health status of the hydraulic structure within the specified period is determined based on the difference between the predicted value and the actual monitoring value of the corresponding data.

2. A method for safety evaluation of an engineering safety monitoring system according to claim 1, characterized in that: In the data collection step, pressure sensors, water level sensors, and temperature sensors are specifically used and arranged at key locations of hydraulic structures. The parameters corresponding to the project seepage pressure, water level, and temperature are then monitored in real time, and seepage pressure data sequences, water level data sequences, and temperature data sequences are obtained.

3. A method for safety evaluation of an engineering safety monitoring system according to claim 2, characterized in that: The outlier removal process is as follows: In the specified period, obtain the seepage pressure data series, water level data series, and temperature data series, and mark them as P i , H i and T i , i = 1, 2, ... n, n represents the number of acquisition points in a specified period; Then calculate P i , H i and T i The corresponding average values ​​are recorded as PP, HP and TP respectively; At the same time, calculate P i , H i and T i The corresponding standard deviations are recorded as PC, HC and TC respectively; Then |P i -PP| is compared with q times PC, and |H i -HP| is compared with q times HC, and |T i -TP|Compare with q times TC: If |P i -PP|>q×PC, then the corresponding P i is an abnormal value, and then obtains i The adjacent P i -1 and P i +1, then calculate P i -1 and P i +1, and take it as the outlier P i 's replacement value; If |H i -HP|>q×HC, it means that the corresponding Hi is an abnormal value, and then obtain the value corresponding to H i H i -1 and H i +1, then calculate H i -1 and H i +1, and use it as the replacement value of the outlier Hi; If |T i -TP|>q×TC, then the corresponding T i is an abnormal value, and then obtains the value corresponding to T i T i -1 and T i +1, then calculate T i -1 and T i +1, and take it as the outlier T i 's replacement value; If |P i -PP|≤q×PC, then determine the corresponding P i The value is normal; If |H i -HP|≤q×HC, then determine the corresponding H i The value is normal; If |Ti-TP|≤q×TC, then determine the corresponding T i The value is normal; Wherein, q is a preset multiple value.

4. A method for safety evaluation of an engineering safety monitoring system according to claim 3, characterized in that: When the i value corresponding to the outlier in the corresponding data sequence is 1 or n, the average value in the corresponding data sequence is used as the replacement value of the outlier in the corresponding data sequence.

5. The method for safety evaluation of an engineering safety monitoring system according to claim 1, characterized in that: The monitoring statistical model is constructed as follows: StepA1. In the historical monitoring cycle, extract the corresponding historical monitoring data; Historical monitoring data includes: Environmental quantities, including but not limited to temperature and water level; Monitoring effect sizes, including but not limited to osmolarity; StepA2, then use the environmental quantity as the independent variable and the monitoring effect quantity as the dependent variable, and use mathematical and statistical analysis methods to establish a regression equation between the two.

6. A method for safety evaluation of an engineering safety monitoring system according to claim 5, characterized in that: in, The regression equation is the core component of the monitoring statistical model, which represents the quantitative relationship between independent variables and dependent variables.

7. A method for safety evaluation of an engineering safety monitoring system according to claim 5, characterized in that: The regression equation is established as follows: StepA2.

1. Obtain multiple sample values ​​corresponding to each independent variable and mark them as X tk , where t = 1, 2, ... e, k = 1, 2, ... r, e represents the number of independent variables, and r represents the number of multiple sample values ​​in each independent variable; At the same time, multiple sample values ​​corresponding to each dependent variable are obtained and marked as Y hk , where h = 1, 2, ... w, k = 1, 2, ... r, h represents the number of dependent variables, and r represents the number of multiple sample values ​​in each dependent variable; StepA2.2, select one independent variable from multiple independent variables, select one dependent variable from multiple dependent variables, and calculate the correlation coefficient between the independent variable and the dependent variable respectively; The calculation formula of the correlation coefficient is as follows: Calculate the independent variable Xa k and the dependent variable Yb k The correlation coefficient R ca ; In the formula, Xa k ∈X tk , Yb k ∈Y hk ; XPa for all Xa k The average value of YPc is the average value of all Yc k The average value of By analogy, the correlation coefficients between all independent variables and dependent variables are calculated, and the correlation coefficient matrix is ​​obtained; StepA2.3, among multiple independent variables, calculate the correlation coefficient between two independent variables respectively; The calculation formula of the correlation coefficient is as follows: Calculate the independent variable Xa1 k and Xa2 k The correlation coefficient R a1a2 ; In the formula, Xa1 k ∈X tk , Xa2 k ∈X tk , and Xa1 k ≠Xa2 k ; XPa1 for all Xa1 k The average value of XPa2 is the average value of all Xa2 k The average value of StepA2.

4. Find the independent variable corresponding to the correlation coefficient with the largest absolute value from the correlation coefficient matrix obtained in StepA2.2, and use it as the first factor introduced into the regression equation; And after the first factor is introduced, the partial correlation coefficients between the remaining independent variables and the dependent variable in the presence of the introduced factor are calculated; The partial correlation coefficient is calculated as follows: In the formula, R c(a1a2) is the partial correlation coefficient, R ca1 ∈R ca , R ca2 ∈R ca , and R ca1 ≠R ca2 ; StepA2.5, further determine whether to select the factor with the greatest correlation with the dependent variable from the remaining factors and introduce it into the regression equation based on the size of the partial correlation coefficient; If the partial correlation coefficient is significant, it is introduced into the equation; If it is not significant, it will not be introduced temporarily, and the relationship between other independent variables and dependent variables will continue to be considered.

8. A method for safety evaluation of an engineering safety monitoring system according to claim 7, characterized in that: in, Whether the partial correlation coefficient is significant is determined by a significance test, and the significance test method is as follows: It is performed on a per introduced factor basis; StepP1, through: Calculate the regression sum of squares SSR; Among them, Y0 h is the mean of all sample values ​​of the dependent variable, Yy hk is the predicted value calculated from the regression equation, and the method used to obtain this predicted value includes the introduced independent variables; Step P2, through: Calculate the regression sum of squares SSR; Among them, Y hk is the actual observed value; Step P3, through: SST = SSR + SSE Calculate the total sum of squares; Step P4, through: Calculate the test statistic F; Among them, u is the number of independent variables introduced; StepP5, extract the preset significance level α; Then compare F with Fα(k,mk-1): If F>Fα(k,mk-1), the factor is significant and is retained in the regression equation; If F≤Fα(k, mk-1), the factor is not significant and is eliminated, and the regression equation is recalculated until all introduced factors pass the significance test.

9. A method for safety evaluation of an engineering safety monitoring system according to claim 8, characterized in that: in, α is a probability value pre-set in hypothesis testing, which is used to determine whether the evidence provided by the sample data is sufficient to reject the null hypothesis. Fα(k, mk-1) is the critical value of the F distribution, which is a specified value calculated based on the distribution table or through statistical software.

10. A method for safety evaluation of an engineering safety monitoring system according to claim 1, characterized in that: The health assessment is as follows: Step G1, according to the establishment method of regression equation, construct the regression equation related to seepage pressure, water level and temperature; The regression equation is expressed as follows: P=b0+b1×H+b2×T+b3×H 2 +b4×T 2 +b5×H×T; Among them, P is the osmotic pressure, H is the water level, T is the temperature, b0, b1, b2, b3, b4, b5 are regression coefficients, and the regression coefficients are determined by regression analysis of historical monitoring data; Step G2, substitute the average values ​​of the water level data series and temperature data series monitored in real time within the specified period into the regression equation, and calculate the predicted value of seepage pressure; It is calculated as follows: Py=b0+b1×HP+b2×TP+b3×HP 2 +b4×TP 2 +b5×HP×TP; Among them, Py is the predicted value of seepage pressure, H is the average value of water level data series, and T is the average value of temperature data series; Step G3, taking the average value of the osmotic pressure data sequence within the specified period as the actual monitored osmotic pressure value; Step G4, calculate the error wc between the predicted osmotic pressure value and the actual monitored osmotic pressure value by: wc = PP-Py; Then the absolute value of its error is compared with the corresponding preset error threshold wcy: If |wc|≤wcy, the health status of the hydraulic structure is considered normal within the specified period; If |wc|>wcy, it is determined that the health status of the hydraulic structure within the specified period is abnormal, that is, there are safety hazards in the hydraulic structure within the specified period, and an early warning signal is generated.

Images