Nine-small-place risk identification and intelligent evaluation method based on artificial intelligence

Abstract

The invention relates to the field of risk assessment, in particular to a nine-small-place risk identification and intelligent assessment method based on artificial intelligence. The method comprises the following steps: firstly, obtaining an evaluation weight of each risk according to the number of occurrence records of each risk of a target place and rectification duration of each risk after occurrence each time, and obtaining a dynamic evaluation value of each risk according to the change of the number of occurrence records of the same risk of all places at the same time point; according to the difference of monitoring data between the occurrence period and the non-occurrence period of each risk of each sensor of the target place, the evaluation weight of each risk of the target place is adjusted, the adjusted evaluation weight of each risk at the current moment is obtained, and then the occurrence possibility of each risk in the target place at the current moment is analyzed; and carrying out real-time early warning on each risk of the target place based on the occurrence possibility. According to the invention, the accuracy of risk prediction of nine small places can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of risk assessment, in particular to a nine small place risk identification and intelligent evaluation method based on artificial intelligence. BACKGROUND

[0002] Nine small places refer to a general term of primary schools or kindergartens, small hospitals, small shops, small catering places, small hotels, small singing and dancing entertainment places, small Internet bars, small beauty and bath places and small production and processing enterprises. These places have relatively serious safety hazards due to weak safety awareness and inadequate safety measures. Therefore, accurate identification and evaluation of various risks existing in nine small places can avoid further expansion of risks and ensure the safety of nine small places.

[0003] In related technologies, sensors are usually used to monitor various types of places, and the data monitored are used to evaluate and predict the risks that may occur in the places by combining fixed weights. However, there are usually a large number of different risks in specific places, each risk has different characteristics in terms of time variation, such as static risks with relatively stable occurrence frequency over time and dynamic risks with unstable occurrence frequency over time. The weights of different types of risks in the evaluation process should be different over time, so that the existing method using fixed weights cannot accurately predict the occurrence of risks in various types of places. SUMMARY

[0004] In order to solve the technical problem that the existing method using fixed weights cannot accurately predict the occurrence of risks in various types of places, the purpose of the present application is to provide a nine small place risk identification and intelligent evaluation method based on artificial intelligence, and the technical solution adopted is as follows: The present application provides a nine small place risk identification and intelligent evaluation method based on artificial intelligence, which comprises: Obtaining the occurrence records of each risk in each type of place at each time point, the rectification time of each risk after each occurrence, and the monitoring data of each sensor in each type of place at each time point in real time; Taking any type of place as a target place, obtaining the evaluation weight of each risk of the target place according to the number of occurrence records of each risk of the target place and the rectification time of each risk after each occurrence, and obtaining the dynamic evaluation value of each risk according to the change of the number of occurrence records of the same risk of all places at the same time point; According to the difference between the monitoring data of each sensor of the target site during the occurrence of each risk and during the non-occurrence of each risk, an association degree between each sensor of the target site and each risk is obtained; according to the difference between the monitoring data of each sensor of the target site at the current moment and during the occurrence of each risk, the association degree between each sensor and each risk, and the dynamic evaluation value of each risk, a weight adjustment coefficient of each risk of the target site at the current moment is obtained; based on the weight adjustment coefficient, the evaluation weight of each risk of the target site is adjusted to obtain an adjusted evaluation weight of each risk of the target site at the current moment; According to the difference between the monitoring data of each sensor of the target site at the current moment and the standard safety threshold, the association degree between each sensor and each risk, and the adjusted evaluation weight of each risk of the target site at the current moment, an occurrence probability of each risk in the target site at the current moment is obtained; based on the occurrence probability, a real-time early warning is given to each risk of the target site.

[0005] Further, the obtaining of the evaluation weight of each risk of the target site comprises: In the target site, an average value of the rectification time lengths after all occurrences of each risk is taken as a whole rectification time length of each risk of the target site; After the number of occurrence records of each risk of the target site and the whole rectification time length are comprehensively processed, an evaluation weight of each risk of the target site is obtained.

[0006] Further, the obtaining of the dynamic evaluation value of each risk comprises: The number of occurrence records of the same risk of all sites at the same time point is taken as the occurrence frequency of each risk at each time point; The discrete degree of the occurrence frequency of each risk at all time points is analyzed to obtain a dynamic evaluation value of each risk.

[0007] Further, the obtaining of the association degree between each sensor of the target site and each risk comprises: An average value of the monitoring data of each sensor of the target site at all moments during the occurrence of each risk is taken as first overall monitoring data of each sensor of the target site with respect to each risk; An average value of the monitoring data of each sensor of the target site at all moments during the non-occurrence of each risk is taken as second overall monitoring data of each sensor of the target site with respect to each risk; The absolute value of the difference between the first overall monitoring data and the second overall monitoring data is normalized to obtain an association degree between each sensor of the target site and each risk.

[0008] Further, the obtaining of the weight adjustment coefficient of each risk of the target site at the current time comprises: taking an absolute value of a difference between the monitoring data of each sensor of the target site at the current time and the second overall monitoring data as an occurrence evaluation value of each sensor of the target site at the current time with respect to each risk; performing a weighted summation on the occurrence evaluation value of each sensor of the target site at the current time with respect to each risk by using the correlation degree between each sensor of the target site and each risk to obtain an initial weight adjustment factor of each risk of the target site at the current time; obtaining the weight adjustment coefficient of each risk of the target site at the current time according to the initial weight adjustment factor of each risk of the target site at the current time and the dynamic evaluation value of each risk.

[0009] Further, the obtaining of the weight adjustment coefficient of each risk of the target site at the current time according to the initial weight adjustment factor of each risk of the target site at the current time and the dynamic evaluation value of each risk comprises: performing a comprehensive processing on the initial weight adjustment factor of each risk of the target site at the current time and the dynamic evaluation value of each risk and then performing a normalization processing to obtain the weight adjustment coefficient of each risk of the target site at the current time.

[0010] Further, the obtaining of the adjustment evaluation weight of each risk of the target site at the current time comprises: taking a product value of the weight adjustment coefficient of each risk of the target site at the current time and the evaluation weight of each risk of the target site as a weight adjustment amount of each risk of the target site at the current time; performing a normalization processing on a sum value of the evaluation weight of each risk of the target site and the weight adjustment amount of each risk of the target site at the current time to obtain the adjustment evaluation weight of each risk of the target site at the current time.

[0011] Further, the obtaining of the occurrence possibility of each risk of the target site at the current time comprises: performing a normalization processing on a difference between the monitoring data of each sensor of the target site at the current time and a standard safety threshold value of each sensor to obtain a risk occurrence factor of each sensor of the target site at the current time; performing a weighted summation on the risk occurrence factor of each sensor of the target site at the current time by using the correlation degree between each sensor of the target site and each risk to obtain an initial occurrence coefficient of each risk of the target site at the current time; Based on the adjustment evaluation weight of each risk of the target site at the current moment, the initial occurrence coefficient is adjusted to obtain the occurrence possibility of each risk of the target site at the current moment.

[0012] Further, the adjustment of the initial occurrence coefficient based on the adjustment evaluation weight of each risk of the target site at the current moment to obtain the occurrence possibility of each risk of the target site at the current moment comprises: The product value of the adjustment evaluation weight of each risk of the target site at the current moment and the initial occurrence coefficient is taken as the occurrence adjustment amount of each risk of the target site at the current moment. The sum value of the initial occurrence coefficient and the occurrence adjustment amount of each risk of the target site at the current moment is normalized to obtain the occurrence possibility of each risk of the target site at the current moment.

[0013] Further, the real-time early warning of each risk of the target site comprises: If the occurrence possibility of each risk of the target site at the current moment is greater than a preset possibility threshold, each risk of the target site is warned.

[0014] The present application has the following beneficial effects: The present application considers that the existing method cannot accurately predict the occurrence of risks in various sites using fixed weights. Firstly, the evaluation weight obtained reflects the reference of the real occurrence of each risk of the target site. At the same time, it is considered that there are not only static risks with stable occurrence frequency changes over time in the target site, but also dynamic risks with unstable occurrence frequency changes over time. In the process of predicting the occurrence of various risks of the target site, the weight of different types of risks should change over time. Therefore, the dynamic evaluation value obtained reflects the degree of change of the occurrence frequency of each risk over time. Considering that the monitoring data of different sensors exhibit different characteristics when a certain specific risk occurs, the correlation between the monitoring data of each sensor and the occurrence of each risk can be reflected by the correlation obtained. Further, the evaluation weight of each risk of the target site is adjusted by the real-time obtained weight adjustment coefficient. The evaluation weight is adjusted to reflect the reference of the real occurrence of each risk of the target site in real time and dynamically. Then, the occurrence possibility of each risk of the target site at the current moment is accurately analyzed, and each risk of the target site is accurately warned, thereby improving the accuracy of risk prediction. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, below will briefly introduce the drawings needed in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flow chart of a nine-small-place risk identification and intelligent evaluation method based on artificial intelligence is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0017] In order to further illustrate the technical means and effects taken by the present application to achieve the predetermined purpose, the specific embodiments, structure, features and effects of a nine-small-place risk identification and intelligent evaluation method based on artificial intelligence according to the present application are described in detail below in combination with the drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0019] The specific scheme of the nine-small-place risk identification and intelligent evaluation method based on artificial intelligence provided by the present application is specifically described below in combination with the drawings.

[0020] Please refer to Figure 1 which shows a flow chart of a nine-small-place risk identification and intelligent evaluation method based on artificial intelligence provided by an embodiment of the present application. The method comprises: Step S1: obtaining the occurrence record of each risk in each type of place at each time point, the rectification time length after each occurrence of each risk, and real-time obtaining the monitoring data of each sensor in each type of place at each time point.

[0021] There are a large number of different risk hidden dangers in nine-small-places, for example, including fire access occupation, insufficient fire-fighting equipment and line aging damage, etc. The embodiment of the present application first collects the safety hidden danger statistical table of each type of place, and obtains the occurrence record of each risk in each type of place at each time point, the rectification time length after each occurrence of each risk from the safety hidden danger statistical table.

[0022] Meanwhile, the embodiment of the present application also needs to install various sensors in each type of site, the installed sensors including, for example, smoke sensors, temperature sensors, carbon monoxide sensors, and electrical fire monitoring detectors, etc., and obtain the monitoring data of each type of sensor in each type of site at each time.

[0023] It should be noted that the dimensions of different types of data are different, and therefore the embodiment of the present application also needs to perform standardization processing on the collected different types of data to eliminate the influence of the dimensions, wherein the data standardization is a technical means well known to those skilled in the art, and will not be described here.

[0024] Step S2: taking any one type of site as a target site, obtaining the evaluation weight of each risk of the target site according to the number of occurrence records of each risk of the target site and the rectification duration after each occurrence of each risk; and obtaining the dynamic evaluation value of each risk according to the change of the number of occurrence records of the same risk of all sites at the same time point.

[0025] In the subsequent steps, the possibility of occurrence of each risk in each type of site needs to be predicted based on the monitoring data of the sensors, but the reference of the real occurrence of each risk represented by the monitoring data of the sensors is different in the prediction process, for example, the possibility of real occurrence of part of the risks is small, and the possibility of real occurrence of another part of the risks is large, therefore, the embodiment of the present application first analyzes any one type of site, takes any one type of site as a target site, the more the number of occurrence records of each risk in the target site, and the longer the rectification duration after each occurrence of each risk in the target site, the greater the reference of the real occurrence of each risk, and therefore the evaluation weight of each risk of the target site can be obtained according to the number of occurrence records of each risk of the target site and the rectification duration after each occurrence of each risk, and the reference of the real occurrence of each risk of the target site is reflected by the evaluation weight.

[0026] Preferably, in one embodiment of the present application, the method for obtaining the evaluation weight of each risk of the target site specifically comprises: In the target site, the average value of the rectification duration after all occurrences of each risk is taken as the overall rectification duration of each risk of the target site, and the number of occurrence records and the overall rectification duration of each risk of the target site are comprehensively processed and normalized, and the calculation result is limited to , so as to obtain the evaluation weight of each risk of the target site.

[0027] In the embodiments of the present application, the combination of the two can be realized by calculating the sum or product value of the number of occurrence records and the overall rectification duration of each risk of the target site, which is not limited herein, and the subsequent step of the combination of two or more data can also be realized by using the same method.

[0028] In one embodiment of the present application, the normalization process can be specifically, for example, a max-min normalization process, and the normalization in the subsequent step can also be a max-min normalization process, and in other embodiments of the present application, other normalization methods can be selected according to the specific range of values or the normalization process can be realized by using an activation function and a hyperbolic tangent function, which will not be repeated and limited herein.

[0029] As an example, in one embodiment of the present application, the expression of the evaluation weight of each risk of the target site can be specifically, for example: wherein, represents the evaluation weight of the first risk of the target site; represents the number of occurrence records of the first risk of the target site; represents the average value of the rectification duration of the first risk in all occurrences in the target site, i.e., the overall rectification duration of the first risk; represents a normalization function used for normalization.

[0030] Since there are not only static risks with stable occurrence frequency changes over time in the target site, but also dynamic risks with unstable occurrence frequency changes over time, the weight of different types of risks should change over time in the process of predicting the occurrence of various risks of the target site, so as to accurately predict the occurrence of various risks. Therefore, the embodiments of the present application analyze the changes of the number of occurrence records of the same risk of all sites at the same time point, and reflect the degree of change of the occurrence frequency of each risk over time through the obtained dynamic evaluation value. The greater the dynamic evaluation value of a certain risk, the more dynamic the occurrence of the risk over time, which belongs to a dynamic risk. The smaller the value, the more stable the occurrence of the risk over time, which belongs to a static risk. The dynamic evaluation value of each risk can be effectively adjusted based on the dynamic evaluation value of each risk in the subsequent step, thereby improving the accuracy of the early warning of each risk.

[0031] Preferably, in one embodiment of the present application, the method for obtaining the dynamic evaluation value of each risk specifically comprises: The number of records of the occurrence of the same risk at the same time point in all sites is recorded as the occurrence number of each risk at each time point, the dispersion degree of the occurrence number of each risk at all time points is analyzed, and a dynamic evaluation value of each risk is obtained.

[0032] In an embodiment of the present application, the standard deviation of the occurrence number of each risk at all time points is taken as the dynamic evaluation value of each risk, thereby realizing the analysis of the dispersion degree of the occurrence number of each risk at all time points, and in other embodiments of the present application, the variance of the occurrence number of each risk at all time points is also taken as the dynamic evaluation value of each risk, which is not limited herein.

[0033] Step S3: obtaining the correlation degree between each sensor of the target site and each risk according to the difference between the monitoring data of each sensor of the target site during the occurrence of each risk and during the non-occurrence of each risk; obtaining the weight adjustment coefficient of each risk of the target site at the current time according to the difference between the monitoring data of each sensor of the target site at the current time and during the occurrence of each risk, the correlation degree between each sensor and each risk, and the dynamic evaluation value of each risk; and adjusting the evaluation weight of each risk of the target site based on the weight adjustment coefficient to obtain the adjusted evaluation weight of each risk of the target site at the current time.

[0034] Since the embodiments of the present application need to predict and warn the risks in the target site according to the monitoring data of the sensors in the target site in the subsequent steps, and the correlation degree of the monitoring data of different sensors and the risks is different for a specific risk type, for example, for the fire risk, the temperature sensor can more reflect whether there is a fire risk compared to other types of sensors, therefore, the embodiments of the present application also need to analyze the difference between the monitoring data of each sensor of the target site during the occurrence of each risk and during the non-occurrence of each risk, and through the obtained correlation of the monitoring data of each sensor and the occurrence of each risk, the possibility of the occurrence of each risk reflected by each sensor can be accurately analyzed based on the correlation degree between the sensors and the risks in the subsequent steps, thereby improving the accuracy of the risk prediction of the nine small sites.

[0035] Preferably, in an embodiment of the present application, the method for obtaining the correlation degree between each sensor of the target site and each risk specifically comprises: The average value of the monitoring data of each sensor of the target site at all time points during the occurrence of each risk is taken as the first overall monitoring data of each sensor of the target site with respect to each risk.

[0036] The average of the monitoring data of each sensor at the target site at all times during the period when each risk has not occurred is used as the second overall monitoring data of each sensor at the target site with respect to each risk.

[0037] The greater the difference between the monitoring data of a certain sensor during a period when a risk occurs and the monitoring data during a period when the risk does not occur, the stronger the correlation between the sensor and the risk. Therefore, the absolute value of the difference between the first overall monitoring data and the second overall monitoring data can be normalized, limiting the calculation result to within a certain range. Within the scope, the correlation between each sensor and each risk in the target location can be obtained.

[0038] As an example, in one embodiment of the present invention, the expression for the correlation between each sensor and each risk at the target location can be specifically, for example, as follows: in, The first place representing the target location Type of sensor and the first The degree of correlation between these risks; The first place representing the target location The type of sensor about the first The first overall monitoring data for this type of risk; The first place representing the target location The type of sensor about the first The second overall monitoring data for this type of risk; This represents the normalization function, used for normalization processing.

[0039] Because the occurrence of each risk in the nine small venues exhibits different characteristics over time—namely, there are static and dynamic risks—and each risk exhibits different degrees of dynamism over time, different weight values ​​need to be assigned to different types of risks in order to accurately predict the probability of various risks occurring in the nine small venues. Therefore, this embodiment of the invention obtains the weight adjustment coefficient of each risk in the target venue at the current moment based on the difference between the monitoring data of each sensor in the target venue at the current moment and the monitoring data during the occurrence of each risk, the correlation between each sensor and each risk, and the dynamic evaluation value of each risk. The larger the weight adjustment coefficient, the greater the degree of adjustment of the evaluation weight of each risk at the current moment. Subsequently, the evaluation weight of each risk in the target venue can be adjusted based on the weight adjustment coefficient, thereby adjusting the weight value when predicting each risk in real time and improving the accuracy of predicting various risks in the target venue.

[0040] Preferably, in one embodiment of the present application, the method for obtaining the weight adjustment factor of each risk of the target site at the current time specifically comprises: The absolute value of the difference between the monitoring data of each sensor of the target site at the current time and the second overall monitoring data is taken as the occurrence evaluation value of each sensor of the target site at the current time with respect to each risk. Since the second overall monitoring data reflects the data characteristics when the risk exhibited by each sensor does not occur, the greater the occurrence evaluation value of a certain sensor at the current time with respect to a certain risk, the greater the possibility of the occurrence of the risk exhibited by the monitoring data of the sensor at the current time, and thus the greater the degree of adjustment of the evaluation weight of the risk in the subsequent.

[0041] The greater the correlation between a certain sensor and a risk, the stronger the connection between the monitoring data characteristics of the sensor and the risk, and thus the correlation between each sensor of the target site and each risk of the target site can be used to perform a weighted summation on the occurrence evaluation value of each sensor of the target site at the current time with respect to each risk, to obtain an initial weight adjustment factor of each risk of the target site at the current time. The greater the initial adjustment factor, the greater the degree of adjustment of the evaluation weight of each risk at the current time.

[0042] As an example, in one embodiment of the present application, the expression of the initial weight adjustment factor of each risk of the target site at the current time can be specifically, for example: wherein, represents the initial weight adjustment factor of the i-th risk of the target site at the current time; represents the correlation between the i-th sensor of the target site and the j-th risk; represents the monitoring data of the i-th sensor of the target site at the current time; represents the second overall monitoring data of the i-th sensor of the target site with respect to the j-th risk; represents the number of sensors in the target site; represents the occurrence evaluation value of the i-th sensor of the target site at the current time with respect to the j-th risk.

[0043] ​​​​​​​​Meanwhile, the greater the dynamic evaluation value of a certain risk, the stronger the dynamic change characteristic of the occurrence of the risk in time, and thus the greater the degree of real-time adjustment of the weight of the risk at the time of prediction, and thus the weight adjustment coefficient of each risk of the target site at the current time can be obtained according to the initial weight adjustment factor of each risk of the target site at the current time and the dynamic evaluation value of each risk.

[0044] Preferably, in an embodiment of the present application, the method for obtaining the weight adjustment coefficient of each risk of the target site at the current time further comprises: comprehending the initial weight adjustment factor of each risk of the target site at the current time and the dynamic evaluation value of each risk and performing normalization processing, limiting the calculation result in the range of 0 to 1, so as to obtain the weight adjustment coefficient of each risk of the target site at the current time.

[0045] As an example, in an embodiment of the present application, the expression of the weight adjustment coefficient of each risk of the target site at the current time can be specifically, for example: wherein, represents the weight adjustment coefficient of the i th risk of the target site at the current time; represents the initial weight adjustment factor of the i th risk of the target site at the current time; represents the initial weight adjustment factor of the i th risk of the target site at the current time; represents the dynamic evaluation value of the i th risk; represents a normalization function for normalization processing. Since the above-mentioned evaluation weight is fixed for a specific risk, and since there are some dynamic risks that change over time in the nine small sites, the prediction of various risks in the nine small sites will be reduced in accuracy when the evaluation weight of each risk is used, and thus the evaluation weight of each risk of the target site can be adjusted based on the weight adjustment coefficient of each risk of the target site at the current time, to obtain the adjusted evaluation weight of each risk of the target site at the current time, and the possibility of the occurrence of each risk of the target site at the current time can be accurately predicted by using the adjusted evaluation weight.

[0046] Preferably, in an embodiment of the present application, the method for obtaining the adjusted evaluation weight of each risk of the target site at the current time specifically comprises:

[0047] Preferably, in an embodiment of the present application, the method for obtaining the adjusted evaluation weight of each risk of the target site at the current time specifically comprises: ​​The product of the weight adjustment factor for each risk in the target location at the current moment and the assessment weight for each risk in the target location is used as the weight adjustment amount for each risk in the target location at the current moment. The sum of the assessment weight and the weight adjustment amount for each risk in the target location at the current moment is then normalized, limiting the calculation result to... Within this range, the adjusted assessment weight of each risk at the target location at the current moment is obtained.

[0048] As an example, in one embodiment of the present invention, the expression for adjusting the assessment weight of each risk of the target location at the current moment can be specifically as follows: in, The first place representing the target location Adjusting the assessment weight of each risk at the current moment; The first place representing the target location The assessment weight of each risk at the current moment; The first place representing the target location The weighting adjustment factor for each type of risk at the current moment; This represents the normalization function, used for normalization processing.

[0049] Step S4: Based on the differences between the monitoring data of each sensor in the target location at the current moment and the standard safety threshold, the correlation between each sensor and each risk, and the adjusted evaluation weight of each risk in the target location at the current moment, obtain the probability of occurrence of each risk in the target location at the current moment; based on the probability of occurrence, provide real-time early warning for each risk in the target location.

[0050] After obtaining the adjusted assessment weight of each risk in the target location at the current moment, the probability of each risk occurring can be predicted by combining the adjusted assessment weight. Since each sensor has a known standard safety threshold, when the data monitored by the sensor exceeds the standard safety threshold, it indicates that a certain risk may occur. Therefore, the probability of each risk occurring in the target location at the current moment can be obtained based on the difference between the monitoring data of each sensor in the target location at the current moment and the standard safety threshold, the correlation between each sensor and each risk, and the adjusted assessment weight of each risk in the target location at the current moment.

[0051] Preferably, in one embodiment of the present invention, the method for obtaining the probability of occurrence of each risk in the target location at the current moment specifically includes: The difference between the monitoring data of each sensor at the target location at the current moment and the standard safety threshold of each sensor is normalized, and the calculation result is limited to... The risk occurrence factor of each sensor of the target site at the current time is obtained within the range, and the greater the risk occurrence factor of a certain sensor at the current time, the greater the possibility of the risk occurrence of the data characteristic of the sensor at the current time.

[0052] The risk occurrence factor of each sensor of the target site at the current time is weighted and summed by using the correlation degree between each sensor of the target site and each risk, and the initial occurrence coefficient of each risk of the target site at the current time is obtained. The greater the initial occurrence coefficient of a certain risk at the current time, the greater the possibility of the risk occurrence at the current time.

[0053] As an example, in an embodiment of the present application, the expression of the initial occurrence coefficient of each risk of the target site at the current time can be specifically, for example: wherein, represents the initial occurrence coefficient of the jth risk of the target site at the current time; represents the correlation degree between the ith sensor and the jth risk of the target site; represents the risk occurrence factor of the ith sensor of the target site at the current time; represents the number of sensors in the target site. Meanwhile, the greater the adjustment evaluation weight of a certain risk at the current time, the greater the possibility of the real risk occurrence, and therefore the initial occurrence coefficient can be adjusted based on the adjustment evaluation weight of each risk of the target site at the current time to obtain the occurrence possibility of each risk of the target site at the current time. Preferably, in an embodiment of the present application, the method for obtaining the occurrence possibility of each risk of the target site at the current time further comprises: The product value of the adjustment evaluation weight and the initial occurrence coefficient of each risk of the target site at the current time is taken as the occurrence adjustment amount of each risk of the target site at the current time, the sum of the initial occurrence coefficient and the occurrence adjustment amount of each risk of the target site at the current time is normalized, and the calculation result is limited within the range, thereby obtaining the occurrence possibility of each risk of the target site at the current time.

[0054]

[0055] As an example, in an embodiment of the present application, the expression of the occurrence possibility of each risk of the target site at the current time can be specifically, for example:

[0056] As an example, in an embodiment of the present application, the expression of the occurrence possibility of each risk of the target site at the current time can be specifically, for example: ​​​​ wherein, represents the occurrence possibility of the i-th risk in the target site at the current time; represents the initial occurrence coefficient of the i-th risk in the target site at the current time; represents the initial occurrence coefficient of the i-th risk in the target site at the current time; represents the initial occurrence coefficient of the i-th risk in the target site at the current time; represents the adjustment evaluation weight of the i-th risk in the target site at the current time; represents the adjustment evaluation weight of the i-th risk in the target site at the current time; represents a normalization function for normalization processing.

[0057] Further, based on the occurrence possibility, real-time early warning is performed on each risk in the target site, so as to improve the accuracy of risk prediction and early warning in the target site.

[0058] Preferably, in an embodiment of the present application, the method of performing real-time early warning on each risk in the target site specifically comprises: if the occurrence possibility of each risk in the target site at the current time is greater than a preset possibility threshold, early warning is performed on each risk in the target site, wherein the preset possibility threshold has a value range of 0 < the preset possibility threshold < 1. In an embodiment of the present application, the preset possibility threshold is set to 0.7, and the specific value of the preset possibility threshold can also be set by the implementer according to the specific implementation scene, which is not limited herein.

[0059] Thus, the above-mentioned same method can be used to realize more accurate risk early warning in various sites in the nine small sites.

[0060] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or can be advantageous.

[0061] Each embodiment in the specification is described in a progressive manner, and the same and similar parts between each embodiment can be referred to each other, and each embodiment mainly describes the differences from other embodiments.

Claims

1. An artificial intelligence-based risk identification and intelligent assessment method for nine small places, characterized in that, The method comprises: acquiring occurrence records of each risk in each type of site at each time point, rectification time of each risk after each occurrence, and monitoring data of each sensor in each type of site at each time point in real time; taking any type of site as a target site, obtaining an evaluation weight of each risk of the target site according to the number of occurrence records of each risk of the target site and the rectification time of each risk after each occurrence, and obtaining a dynamic evaluation value of each risk according to changes in the number of occurrence records of the same risk of all sites at the same time point; obtaining a correlation degree between each sensor of the target site and each risk according to differences in the monitoring data of each sensor of the target site between the risk occurrence period and the non-risk occurrence period, obtaining a weight adjustment coefficient of each risk of the target site at the current time according to differences in the monitoring data of each sensor of the target site between the current time and the risk occurrence period, the correlation degree between each sensor and each risk, and the dynamic evaluation value of each risk, and adjusting the evaluation weight of each risk of the target site based on the weight adjustment coefficient to obtain an adjusted evaluation weight of each risk of the target site at the current time; obtaining a possibility of occurrence of each risk in the target site at the current time according to differences between the monitoring data of each sensor of the target site at the current time and a standard safety threshold, the correlation degree between each sensor and each risk, and the adjusted evaluation weight of each risk of the target site at the current time, and performing real-time early warning on each risk of the target site based on the possibility of occurrence.

2. The method of claim 1, wherein the method is characterized by, The method comprises: in the target site, taking an average value of the rectification time of each risk after all occurrences as an overall rectification time of each risk of the target site; comprehensively processing the number of occurrence records of each risk of the target site and the overall rectification time and performing normalization processing to obtain the evaluation weight of each risk of the target site.

3. The method of claim 1, wherein the method further comprises: determining a risk level of each of the nine small places based on the risk factors; and providing a risk assessment report of each of the nine small places based on the risk level of each of the nine small places. The method comprises: taking the number of occurrences of each risk at each time point as the number of occurrences of each risk at each time point; analyzing the discrete degree of the number of occurrences of each risk at all time points to obtain the dynamic evaluation value of each risk.

4. The method of claim 1, wherein the method is characterized by, The method comprises: taking an average value of the monitoring data of each sensor of the target site at all time points during the occurrence of each risk as first overall monitoring data of each sensor of the target site with respect to each risk; taking an average value of the monitoring data of each sensor of the target site at all time points during the non-occurrence of each risk as second overall monitoring data of each sensor of the target site with respect to each risk; performing normalization processing on the absolute value of the difference between the first overall monitoring data and the second overall monitoring data to obtain the correlation degree between each sensor of the target site and each risk.

5. The method for risk identification and intelligent assessment of small venues based on artificial intelligence according to claim 4, characterized in that, The obtaining of the weight adjustment coefficient of each risk of the target site at the current moment comprises: taking the absolute value of the difference between the monitoring data of each sensor of the target site at the current moment and the second overall monitoring data as the occurrence evaluation value of each sensor of the target site at the current moment with respect to each risk; performing weighted summation on the occurrence evaluation value of each sensor of the target site at the current moment with respect to each risk by using the correlation degree between each sensor of the target site and each risk to obtain an initial weight adjustment factor of each risk of the target site at the current moment; obtaining the weight adjustment coefficient of each risk of the target site at the current moment according to the initial weight adjustment factor of each risk of the target site at the current moment and the dynamic evaluation value of each risk.

6. The method according to claim 5, wherein the method is characterized by, The obtaining of the weight adjustment coefficient of each risk of the target site at the current moment according to the initial weight adjustment factor of each risk of the target site at the current moment and the dynamic evaluation value of each risk comprises: performing comprehensive processing on the initial weight adjustment factor of each risk of the target site at the current moment and the dynamic evaluation value of each risk and then performing normalization processing to obtain the weight adjustment coefficient of each risk of the target site at the current moment.

7. The method according to claim 1, wherein the method is characterized by, The obtaining of the adjustment evaluation weight of each risk of the target site at the current moment comprises: taking the product value of the weight adjustment coefficient of each risk of the target site at the current moment and the evaluation weight of each risk of the target site as the weight adjustment amount of each risk of the target site at the current moment; performing normalization processing on the sum of the evaluation weight of each risk of the target site and the weight adjustment amount of each risk of the target site at the current moment to obtain the adjustment evaluation weight of each risk of the target site at the current moment.

8. The method according to claim 1, wherein the method is characterized by, The obtaining of the occurrence possibility of each risk of the target site at the current moment comprises: performing normalization processing on the difference between the monitoring data of each sensor of the target site at the current moment and the standard safety threshold value of each sensor to obtain a risk occurrence factor of each sensor of the target site at the current moment; performing weighted summation on the risk occurrence factor of each sensor of the target site at the current moment by using the correlation degree between each sensor of the target site and each risk to obtain an initial occurrence coefficient of each risk of the target site at the current moment; adjusting the initial occurrence coefficient based on the adjustment evaluation weight of each risk of the target site at the current moment to obtain the occurrence possibility of each risk of the target site at the current moment.

9. The method according to claim 8, wherein the nine small places risk identification and intelligent assessment based on artificial intelligence is characterized in that, The adjusting of the initial occurrence coefficient based on the adjustment evaluation weight of each risk of the target site at the current moment to obtain the occurrence possibility of each risk of the target site at the current moment comprises: taking the product value of the adjustment evaluation weight of each risk of the target site at the current moment and the initial occurrence coefficient as the occurrence adjustment amount of each risk of the target site at the current moment. The sum of the initial occurrence coefficient and the occurrence adjustment amount of each risk of the target site at the current time is normalized to obtain the occurrence probability of each risk of the target site at the current time.

10. The method of claim 1, wherein the method is based on artificial intelligence. The real-time early warning of each risk of the target site comprises: If the occurrence probability of each risk of the target site at the current time is greater than a preset probability threshold, each risk of the target site is early warned.

Images