Data processing method of intelligent security comprehensive management and control platform

Abstract

The invention relates to the technical field of data processing, in particular to a data processing method of an intelligent security comprehensive management and control platform. The method comprises the following steps: acquiring environment smoke concentration data; the method comprises the following steps: by taking data acquired in real time as real-time data, forming an observation interval by # imgabs0 # data before acquiring the real-time data; determining an oscillation factor of the real-time data according to the fluctuation condition of the data in the observation interval; calculating a noise optimization factor of the real-time data; taking a normalization result of a product of the noise optimization factor and the oscillation factor as a scale adjustment factor; adjusting a preset initial scale difference through the scale adjustment factor to obtain an optimal scale difference; and dividing a plurality of scales based on the optimal scale difference to calculate a multi-scale entropy to monitor the environment. The method has the effect of improving the system calculation efficiency and the monitoring accuracy.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and particularly to a data processing method for an intelligent security comprehensive control platform. Background Art

[0002] The occurrence of a fire will cause losses in life and property. Therefore, how to prevent the occurrence of a fire is an issue that people have been continuously concerned about. Usually, manual inspections can be used to prevent the occurrence of a fire. However, manual inspections have a lag in response and usually cannot detect a fire in a timely manner. With the development of electronic information technology and the requirement for the timeliness of fire prevention, the method of detecting smoke based on sensors is applied to fire prevention. During the early stage and the process of a fire, thick smoke is usually generated. Therefore, by monitoring the environmental smoke concentration with a smoke sensor, the occurrence of a fire can be detected in a timely manner. The currently common method for detecting and analyzing smoke concentration is the multi-scale entropy algorithm, which can capture the multi-scale changes of data and has good adaptability and robustness.

[0003] When calculating the multi-scale entropy value of each data sequence, the multi-scale entropy algorithm usually sets the numerical difference between all adjacent scales to 1 after a preset number of scales. For example, when calculating the multi-scale entropy value of a data sequence, the number of scales is 6, and the sizes of these six scales are 1, 2, 3, 4, 5, and 6 in sequence. The size of the scale affects the subsequent calculation amount and also affects the accuracy of the final data feature extraction. However, for simple signals, too small scales are not required. Therefore, the fixed-scale entropy algorithm in the traditional multi-scale entropy algorithm will bring a certain burden to the calculation, resulting in low calculation efficiency. Long-term high-load operation may lead to a decline in system performance; furthermore, the situation where an early warning cannot be issued in a timely manner when a fire occurs may occur. Summary of the Invention

[0004] To solve the situation where the fixed-scale calculation efficiency in the traditional multi-scale entropy algorithm is low, resulting in a decline in system performance and affecting the detection accuracy, this application provides a data processing method for an intelligent security comprehensive control platform.

[0005] In a first aspect, this application provides a data processing method for an intelligent security comprehensive control platform, adopting the following technical solution: A data processing method for an intelligent security comprehensive control platform includes the steps of: obtaining environmental smoke concentration data; using the real-time collected data as real-time data, and obtaining the data before the real-time data A set of data constitutes an observation interval; determine the oscillation factor of the real-time data according to the fluctuation of the data in the observation interval; calculate the noise optimization factor of the real-time data; use the normalized result of the product of the noise optimization factor and the oscillation factor as the scale adjustment factor; adjust the preset initial scale difference through the scale adjustment factor to obtain the optimal scale difference; divide multiple scales based on the optimal scale difference and calculate the multi-scale entropy to monitor the environment; Among them, the steps of calculating the noise optimization factor of the real-time data include: obtaining the extreme values of the data in the observation interval, calculating the deviation based on the difference between the extreme values and the adjacent data, and taking the mean of multiple deviations as the discrete difference; calculating the absolute value of the difference between two adjacent data in the observation interval, and taking the mean of multiple absolute values as the adjacent difference; taking the normalized result of the absolute difference between the discrete difference and the adjacent difference as the data dispersion; determining the noise optimization factor based on the data dispersion, and the data dispersion is proportional to the noise optimization factor.

[0006] The beneficial effects are as follows: During the process of real-time collecting smoke concentration data, a real-time data observation interval is constructed in real time. Set the oscillation factor according to the fluctuation of the data in the observation interval, adjust the initial scale difference according to the oscillation factor to divide the scale, and monitor the environment for fire according to the multi-scale entropy. During the monitoring process, each real-time data corresponds to a different observation space. Furthermore, the multi-scale division corresponding to each real-time data is different, and as the real-time data is continuously collected, the scale division also changes dynamically. According to the actual data fluctuation situation, the optimal scale is divided for the real-time data. On the one hand, it can improve the calculation efficiency of the multi-scale entropy. On the other hand, the dynamic scale division also improves the accuracy of the final multi-scale entropy calculation, thereby improving the accuracy of environmental monitoring.

[0007] At the same time, the noise optimization factor is also calculated in this method to analyze whether there is noise in the observation interval and reduce the influence of noise on the calculation of the final optimal scale difference, thereby further improving the accuracy of environmental monitoring.

[0008] The characteristics of noise are mainly manifested as being more discrete in spatial distribution, and the change of data is inconsistent with the overall change in the observation interval. Therefore, the difference between the discrete difference and the adjacent difference is calculated to reflect the difference between the data change at the extreme point and the data change in the observation interval, and then it can be judged whether the extreme point is noise.

[0009] Optionally, the steps of determining the oscillation factor of the real-time data according to the fluctuation of the data in the observation interval include: obtaining the data difference between adjacent data in the observation interval, and obtaining the mean of the data differences; taking the mean of the data differences as the fluctuation amplitude; obtaining the extreme point density of the observation interval, and determining the oscillation factor based on the fluctuation amplitude and the extreme point density.

[0010] The beneficial effects are as follows: The data difference between adjacent data represents the degree of change between two adjacent data. Calculate the mean of the data differences to reflect the overall speed of data change in the observation interval, which can also be understood as reflecting whether the data change in the observation interval is gentle. The larger the adjacent difference, the faster the data change and the more violent the fluctuation in the data observation interval. The density of extreme points represents the number of extreme points in the observation interval and can reflect the frequency of change in the observation interval. The two act synergistically on the oscillation factor, thereby improving the accuracy of oscillation factor calculation.

[0011] Optionally, take the product of the fluctuation amplitude and the extreme point density in the observation interval as the fluctuation degree, and determine the oscillation factor of the observation interval. The fluctuation degree is inversely proportional to the oscillation factor.

[0012] Optionally, take the ratio of the number of extreme points in the observation sequence to the length of the observation interval as the extreme point density.

[0013] The beneficial effects are as follows: The ratio of the number of extreme points to the length of the observation interval represents the proportion of extreme points in the data of the observation interval; the larger this proportion, the more times the data fluctuates within a specific time interval, and thus it also indicates that the data in the observation interval fluctuates more violently.

[0014] Optionally, in the step of calculating the deviation based on the difference between the extreme value and adjacent data: In response to the extreme point being at the end of the sequence, take the absolute value of the difference between the extreme point and the data point on its own side as the deviation; in response to data points existing on both sides of the extreme point, take the mean of the absolute differences between the extreme point and the two data points as the deviation.

[0015] The beneficial effects are as follows: There are two cases for extreme points here. The first case is that the extreme point is at the end of the observation interval and there is only one adjacent data. For this type of situation, calculate the deviation based on the difference between the extreme point and one data. The other case is that data points exist on both sides of the extreme point, and then take the mean of the differences between the two data and the extreme point data as the deviation.

[0016] Optionally, in the step of calculating the deviation based on the difference between the extreme value and adjacent data: In response to the extreme point being at the end of the sequence, take the absolute value of the difference between the extreme point and the data point on its own side as the deviation; in response to data points existing on both sides of the extreme point, obtain the absolute differences between the extreme point and the two data points, and take the largest data among the corresponding absolute differences of the two data points as the deviation The beneficial effects are as follows: In this method, when there are two adjacent data points at the extreme point, select the larger data difference as the deviation, which pays more attention to the local data change at the extreme point.

[0017] Optionally, determining the noise optimization factor based on data dispersion includes the steps of: obtaining the sampling time interval of extreme points, and taking the product of the sampling time interval and the data dispersion as the noise optimization factor.

[0018] Optionally, the step of obtaining the sampling time interval of the extreme points includes obtaining the observation interval the largest extreme points, and taking the average value of the time intervals between the extreme points as the sampling time interval.

[0019] Optionally, the number of data points in the observation interval is not less than 300 and not more than 500.

[0020] The beneficial effect is that this is mainly used to control the amount of data in the observation interval. On the one hand, it prevents the amount of data in the observation interval from being too small to reflect the fluctuation situation. On the other hand, it avoids introducing data that is too far from the real-time data.

[0021] Second, the present application provides a data processing system for an intelligent security comprehensive control platform, adopting the following technical solution: A data processing system for an intelligent security comprehensive control platform includes: a processor and a memory, and the memory stores computer program instructions, which when executed by the processor implement a data processing method for an intelligent security comprehensive control platform according to the above.

[0022] The beneficial effect is that the data processing method for an intelligent security comprehensive control platform according to the above is generated into a computer program and stored in the memory to be loaded and executed by the processor. Thus, a system is made according to the memory and the processor, which is convenient to use.

[0023] The present application has the following technical effects: According to the fluctuation situation in the observation interval corresponding to the real-time data, the optimal scale corresponding to the real-time data is dynamically divided. On the one hand, it can reduce unnecessary calculation amounts, and on the other hand, the information of the extracted data is more accurate, thereby improving the accuracy of environmental monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a flowchart of a data processing method for an intelligent security comprehensive control platform according to an embodiment of the present application.

[0025] Figure 2 is a flowchart of step S3 in the method of a data processing method for an intelligent security comprehensive control platform according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0026] The embodiments of the present application disclose a data processing method for an intelligent security comprehensive management and control platform. During the process of detecting the environment, smoke concentration data is collected in real time as real-time data. An observation interval of the real-time data is constructed, the fluctuations of the data within the observation interval are analyzed, and an oscillation factor is determined; a noise optimization factor of the real-time data is obtained. The preset scale difference between different scales is adjusted by the oscillation factor and the noise optimization factor. Thus, the optimal scale difference is obtained, the scale is re-divided based on the optimal scale difference, and sample entropy is calculated to detect the environment. This method divides the observation interval in real time according to the data collected in real time, and the finally obtained optimal scale difference can be adjusted in real time according to the data in the observation interval. The scale is dynamically divided according to the fluctuation of the data, different scales are set for different data, unnecessary computational burdens are reduced, the computational efficiency is improved, and the situation where the system performance degrades and timely warning cannot be given is reduced.

[0027] Referring to Figure 1 , a data processing method for an intelligent security comprehensive management and control platform includes steps S1 - S4.

[0028] S1: Obtain the environmental smoke concentration data; take the data collected in real time as real-time data, and obtain the first data to form an observation interval.

[0029] Collect the smoke concentration data in the environment to be detected (the environment to be detected can be a forest, a shopping mall, etc.). In this embodiment, a smoke sensor is used to collect the smoke concentration data, and the collection frequency is once per second. In other embodiments, the collection frequency can be adjusted.

[0030] As the smoke concentration data accumulates continuously, a smoke data sequence is formed. Define the data collected in real time as real-time data. Take the first data to form an observation interval, which is not less than 300 and not more than 500 to avoid introducing overly old data while maintaining a certain amount of data. In this embodiment is 400, and in other embodiments, it can be adjusted according to the actual situation of the size.

[0031] S2: Determine the oscillation factor of the real-time data according to the fluctuation of the data in the observation interval; Obtain the data differences between adjacent data in the observation interval, and obtain the average value of the data differences; take the average value of the data differences as the fluctuation amplitude; obtain the extreme point density of the observation interval, and determine the oscillation factor based on the fluctuation amplitude and the extreme point density; take the product of the fluctuation amplitude and the extreme point density in the observation interval as the fluctuation degree, determine the oscillation factor of the observation interval, and the fluctuation degree is inversely proportional to the oscillation factor.

[0032] Specifically, the calculation formula of the oscillation factor of the observation interval corresponding to the real-time data can be expressed as: ; In the formula, The oscillation factor representing the observation interval of real-time data; Indicates the fluctuation range of data in the observation interval; Indicates the number of extreme points in the observation interval; Indicates the length of the observation interval; represents a hyperparameter; The natural constant An exponential function with base .

[0033] The larger the oscillation factor, the more violent the fluctuation in the observation interval, and the smaller the oscillation factor, the gentler the oscillation in the observation interval. For real-time data with gentle fluctuations in the observation interval, a larger scale difference is suitable, while for real-time data with violent vibrations in the observation interval, a smaller scale difference is suitable.

[0034] The hyperparameter is a positive number, and the hyperparameter is mainly used to prevent the oscillation factor from being zero. In this embodiment, the hyperparameter is 0.001.

[0035] The absolute difference between adjacent data in the observation interval is the data difference. There is a data difference between any two adjacent data. The mean of the data difference reflects the overall fluctuation degree of the data in the observation interval. Therefore, the mean of the data difference is taken as the fluctuation amplitude.

[0036] The more extreme points there are in an observation interval, the higher the frequency of data fluctuations in a limited time period. Therefore, the ratio of the number of extreme points in the observation sequence to the length of the observation interval is taken as the extreme point density. The extreme point density can reflect the fluctuation frequency of data in the observation interval.

[0037] S3: Calculate the noise optimization factor of real-time data.

[0038] In the calculation process of the oscillation factor, the change characteristics of the data in the observation interval determine the oscillation factor of the corresponding observation interval. However, in the actual data collection process, the sensor detection may be affected by electromagnetic interference or signal loss, which may lead to data fluctuations in the observation interval, thereby affecting the accuracy of the final calculation of the oscillation factor. Therefore, the noise optimization factor is calculated here to adjust and optimize the oscillation factor to reduce the impact of signal noise on the calculation of the oscillation factor.

[0039] Reference Figure 2 The step of calculating the noise optimization factor of real-time data includes: step S31-step S34.

[0040] S31: Obtain the extreme value of the data in the observation interval, calculate the deviation based on the difference between the extreme value and the adjacent data, and take the mean of multiple deviations as the discrete difference; Extract the extreme points (which can also be understood as the common set of peak points and valley points) from the data curve formed by the data within the observation interval. Calculate the absolute value of the difference between the extreme points and the adjacent data, and use this absolute value as the deviation. And take the mean of multiple deviations as the discrete difference of the observation interval. The discrete difference represents the degree of dispersion of the extreme values in the overall data of the observation interval. During the data acquisition process, the difference between noise data and real data usually lies in that the noise data is more dispersed in the data of the observation interval. Therefore, the discrete difference can reflect the possibility that the data in the observation interval contains noise.

[0041] S32: Calculate the absolute value of the difference between two adjacent data in the observation interval, and take the mean of multiple absolute values as the adjacent difference.

[0042] The adjacent difference calculates the average difference between any two data, and can reflect to a certain extent the speed of data change in the observation interval, reflecting the overall change trend of the data in the observation interval. This trend is a normal change.

[0043] S33: Take the normalized result of the absolute difference between the discrete difference and the adjacent difference as the data dispersion degree.

[0044] In this embodiment, through the function normalizes the absolute difference between the discrete difference and the adjacent difference. Comparing the change of data at the normal change and the extreme points, if the magnitude of the discrete difference is close to the magnitude of the adjacent difference, it indicates that the data change trend corresponding to the extreme points is similar to the normal change trend. Then it shows that the change at the extreme points in the observation interval is more likely to be a normal change, and the possibility of noise in the observation interval is smaller.

[0045] S34: Determine the noise optimization factor based on the data dispersion degree. The data dispersion degree is proportional to the noise optimization factor.

[0046] Among all the extreme points, select the extreme point with the largest value, obtain the time interval between adjacent extreme points, and take the mean of multiple time intervals as the sampling time interval.

[0047] In the observation interval, if the largest several numerical points in the observation interval are dispersed, it also indicates the poor gradualness between these numerical points, and further indicates the greater possibility of noise in the observation interval. Therefore, here the product of the sampling time interval and the data dispersion degree is used as the noise optimization factor.

[0048] Specifically, the calculation formula of the noise optimization factor can be expressed as: ; where represents the noise optimization factor of real-time data; represents the sampling time interval; represents the discrete difference of the observation interval; represents the adjacent difference of the data in the observation interval; represents the linear normalization function.

[0049] During the calculation process, if the number of extreme points in the observation interval is 0, the noise optimization factor is 1.

[0050] S4: Use the normalized result of the product of the noise optimization factor and the oscillation factor as the scale adjustment factor; adjust the preset initial scale difference through the scale adjustment factor to obtain the optimal scale difference; divide multiple scales based on the optimal scale difference and calculate the multi-scale entropy to monitor the environment.

[0051] Multiply the scale adjustment factor by the preset scale difference and perform normalization processing to obtain the scale adjustment factor, and multiply the scale adjustment factor by the preset initial scale difference to obtain the optimal scale difference. To facilitate intervention on the optimal scale difference, a adjustment coefficient is set here, and the product of the scale adjustment factor, the oscillation factor, and the adjustment coefficient is used as the optimal scale difference. In this embodiment, the adjustment coefficient is 5. The setting of the adjustment coefficient enables the final optimal scale difference to not only be automatically adjusted according to the data change, but also be intervened manually, further improving the accuracy of detection.

[0052] After the optimal scale difference is calculated, the scale is divided. For example: the set initial scale is 1, the optimal scale difference is 6, and the number of scales is 6, then the multiple scales finally divided are: 1, 7, 13, 19, 25, 31.

[0053] Multiply the noise optimization factor by the oscillation factor to obtain the scale adjustment factor for adjusting the scale. The size of the scale adjustment factor increases as the oscillation factor increases. The greater the fluctuation of the data in the observation interval, the smaller the scale difference should be selected for calculating the sample entropy in the subsequent calculation. For the real-time data corresponding to the observation interval with small fluctuations, a larger scale difference can be used to calculate the sample entropy to improve the recognition and capture effect of anomalies.

[0054] After the scale division is completed, calculate the sample entropy of multiple scales of the real-time data, and use the normalized result of the mean of multiple sample entropies as the multi-scale entropy of the real-time data; Set an anomaly threshold. In response to the multi-scale entropy being greater than the anomaly threshold and the head-end data in the observation interval being less than the tail-end data, an alarm is issued for the smoke concentration.

[0055] The head-end data in the observation interval is less than the tail-end data, indicating that the smoke concentration continues to rise during this time period. At the same time, combined with the multi-scale entropy to judge the fire, further improving the accuracy of fire monitoring.

[0056] The embodiment of the present application also discloses a data processing system for an intelligent security comprehensive management and control platform, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a data processing method for an intelligent security comprehensive management and control platform according to the present application is implemented.

[0057] The above system also includes other components well-known to those skilled in the art, such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be elaborated here.

[0058] The above are all the preferred embodiments of the present application. The protection scope of the present application is not limited thereby. Therefore, all equivalent changes made according to the structure, shape, and principle of the present application shall be covered within the protection scope of the present application.

Claims

1. A data processing method for an intelligent security integrated management and control platform, characterized in that: The steps include: obtaining environmental smoke concentration data; taking the real-time collected data as real-time data, and obtaining the real-time data before Each piece of data constitutes the observation interval; Determine the oscillation factor of real-time data according to the fluctuation of data in the observation interval; calculate the noise optimization factor of real-time data; The normalized result of the product of the noise optimization factor and the oscillation factor is used as the scale adjustment factor; The preset initial scale difference is adjusted by the scale adjustment factor to obtain the optimal scale difference; Based on the optimal scale difference, multiple scales are divided to calculate multi-scale entropy to monitor the environment; Among them, the step of calculating the noise optimization factor of real-time data includes: obtaining the extreme value of the data in the observation interval, calculating the deviation based on the difference between the extreme value and the adjacent data, and taking the mean of multiple deviations as the discrete difference; calculating the absolute value of the difference between two adjacent data in the observation interval, and taking the mean of multiple absolute values ​​as the adjacent difference; taking the normalized result of the absolute difference between the discrete difference and the adjacent difference as the data discreteness; determining the noise optimization factor based on the data discreteness, and the data discreteness is proportional to the noise optimization factor.

2. The data processing method of the intelligent security integrated management and control platform according to claim 1 is characterized in that: The steps of determining the oscillation factor of real-time data according to the fluctuation of data in the observation interval include: obtaining the data difference between adjacent data in the observation interval, obtaining the mean of the data difference; taking the mean of the data difference as the fluctuation amplitude; obtaining the extreme point density of the observation interval, and determining the oscillation factor based on the fluctuation amplitude and the extreme point density.

3. The data processing method of the intelligent security integrated management and control platform according to claim 2 is characterized in that: The product of the fluctuation amplitude and the extreme point density in the observation interval is taken as the volatility, and the oscillation factor of the observation interval is determined. The volatility is inversely proportional to the oscillation factor.

4. The data processing method of the intelligent security integrated management and control platform according to claim 2 is characterized in that: The ratio of the number of extreme points in the observation sequence to the length of the observation interval is taken as the extreme point density.

5. The data processing method of the intelligent security integrated management and control platform according to claim 1 is characterized in that: In the step of calculating the deviation based on the difference between the extreme value and the adjacent data: in response to the extreme value point being located at the end of the sequence, the absolute value of the difference between the extreme value point and the data points on its side is taken as the deviation; in response to the presence of data points on both sides of the extreme value point, the average of the absolute differences between the extreme value point and the two data points is taken as the deviation.

6. The data processing method of the intelligent security integrated management and control platform according to claim 1 is characterized in that: In the step of calculating the deviation based on the difference between the extreme value and the adjacent data: in response to the extreme value point being located at the end of the sequence, the absolute value of the difference between the extreme value point and the data points on its side is taken as the deviation; in response to the presence of data points on both sides of the extreme value point, the absolute difference between the extreme value point and the two data points is obtained, and the largest data difference between the absolute differences corresponding to the two data points is taken as the deviation.

7. The data processing method of the intelligent security integrated management and control platform according to claim 1 is characterized in that: Determining the noise optimization factor based on data discreteness includes the following steps: obtaining the extreme point sampling time interval, and taking the product of the sampling time interval and the data discreteness as the noise optimization factor.

8. The data processing method of the intelligent security integrated management and control platform according to claim 7 is characterized in that: The steps of obtaining the extreme point sampling time interval include obtaining the observation interval The largest extreme point, The mean of the time intervals between extreme points is taken as the sampling time interval.

9. The data processing method of the intelligent security integrated management and control platform according to claim 1 is characterized in that: The number of data points in the observation interval is not less than 300 and not more than 500.

10. A data processing system for an intelligent security integrated management and control platform, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a data processing method for an intelligent security integrated management and control platform according to any one of claims 1 to 9 is implemented.

Images