Dynamic scheduling method for global security and protection resources

Abstract

The invention relates to the technical field of digital data processing, in particular to a global security resource dynamic scheduling method, which comprises the following steps of: acquiring a global security pressure data sequence at a current moment and a historical set time window thereof, and performing smoothing processing on the global security pressure data sequence by using an improved exponential smoothing algorithm to obtain global security pressure data at the next moment, acquiring video response frequency at the next moment based on the global security pressure, and performing automatic scheduling based on the video response frequency. According to the invention, the problem of low overall efficiency and reliability of the security system in a complex environment is solved.

Description

Dynamic scheduling method for security resources across the entire domain Technical Field

[0001] This invention relates to the field of digital data processing technology. More specifically, this invention relates to a method for dynamic scheduling of security resources across the entire domain. Background Technology

[0002] With the rapid development of intelligent security systems, the demand for real-time performance, high precision, and resource optimization is increasing in scenarios such as urban public safety, industrial park monitoring, and large building complex management. Traditional security systems often rely on fixed frame rate video acquisition, uniform bandwidth allocation, and static analysis task scheduling strategies, which cannot flexibly adjust resources according to the dynamic changes in each defense zone. This can easily lead to monitoring delays or analysis lags in some high-risk areas, while resources in other low-risk areas remain idle for extended periods, resulting in low overall security efficiency.

[0003] To address this issue, some existing solutions have introduced exponential smoothing methods based on historical data to smooth and predict security stress or monitoring indicators. Exponential smoothing assigns weights to current observations and combines them with a weighted average of past observations, which can suppress data noise and improve the stability of continuous data sequences to some extent.

[0004] However, existing methods generally employ a fixed smoothing coefficient, meaning that the smoothing coefficient remains constant regardless of the fluctuation amplitude and rate of change in the security situation. This fixed-coefficient approach has significant shortcomings in real-world, comprehensive security scenarios: First, when local or overall security zone pressure changes suddenly, the fixed smoothing coefficient leads to a lag in response to new information, failing to reflect rapid changes in the security situation in a timely manner and potentially missing critical anomalies or alarm triggering opportunities. Second, when the security situation remains stable for a long period, the fixed coefficient cannot fully utilize historical data smoothing sequences and is easily affected by occasional noise, causing unnecessary fluctuations in pressure or monitoring indicators, thus affecting the accurate judgment and decision-making regarding the system status. Furthermore, the fixed smoothing coefficient method lacks adaptability to different zones or time periods, and cannot rationally allocate computing resources, network bandwidth, and analysis task priorities in dynamic, multi-source data environments, resulting in both resource waste and potential monitoring blind spots in actual operation. These problems are particularly evident when facing the need for dynamic scheduling of security resources across the entire domain. This is because a comprehensive security system requires real-time collection, analysis, and response to large amounts of multimodal data, including video frame differences, audio decibels, and crowd density, and dynamically allocates resources based on the pressure on each zone. If the exponential smoothing method with a fixed smoothing coefficient is still used, the system will struggle to achieve intelligent and dynamic resource scheduling across all zones while maintaining real-time performance and accuracy. This results in low overall performance and reliability of the security system in complex environments. Summary of the Invention

[0005] To address the problem of low overall performance and reliability of security systems in complex environments as mentioned in the background art, the present invention provides the following solution.

[0006] In a first aspect, the present invention provides a method for dynamic scheduling of global security resources, comprising: acquiring a global security pressure data sequence at the current moment and within a historical set time window; smoothing the global security pressure data sequence using an improved exponential smoothing algorithm to obtain global security pressure data at the next moment; acquiring the video response frequency at the next moment based on the global security pressure data; and performing automatic scheduling based on the video response frequency; wherein the improved exponential smoothing algorithm includes a smoothing coefficient, the smoothing coefficient being positively correlated with the initial value and the regional situation stability at the previous moment, and negatively correlated with the regional situation stability at the current moment.

[0007] The aforementioned technical solution processes the overall security stress sequence using an improved exponential smoothing algorithm based on current and historical data. This adaptively balances the responsiveness to historical trends and recent changes, generating more stable and accurate predictions for the next moment. Combined with the video response frequency calculated from this stress prediction, monitoring resources can be dynamically adjusted, enabling automatic scheduling of camera acquisition parameters, analysis task priorities, and bandwidth allocation. This significantly improves the real-time performance, prediction accuracy, and resource utilization efficiency of the overall security system, while also enhancing sensitivity to abnormal events and the overall robustness of the system.

[0008] Furthermore, the smoothing coefficient for: , As the initial value, , The first Time, Number The stability of the regional situation at any given time.

[0009] The aforementioned technical solution dynamically adjusts the smoothing coefficient based on the regional situational stability at adjacent time points. This enhances the responsiveness to the latest data when situational fluctuations are significant, while reducing sensitivity to occasional changes when the situation is stable, thus achieving adaptive smoothing of the entire security pressure sequence. This mechanism enables pressure prediction to quickly capture sudden events and abnormal fluctuations while maintaining the stability and continuity of the data sequence, effectively improving the accuracy and real-time performance of security situational analysis. Simultaneously, it enhances the reliability of subsequent video response frequency calculations and automatic resource scheduling, as well as the overall robustness of the system.

[0010] Furthermore, the regional situational stability for: , For the natural constant An exponential function with base 0. For the first Standard deviation of the overall security pressure data under the specified time window and its historical settings. For the first Standard deviation of the overall security pressure data under the given time window and its historical settings. For the first The maximum value among the standard deviations of the overall security stress data for all times within a given time window and its historical settings. For the first The size of the historical time window corresponding to each moment. These are the preset hyperparameters.

[0011] The aforementioned technical solution constructs an exponential regional situational stability model based on the fluctuation characteristics of full-domain security pressure data within a historical time window. This model can reflect changes in the regional situation over time in a continuous and smooth manner. When the current pressure fluctuation is close to historical pressure characteristics, the stability value is high, indicating that the regional state remains consistent. Conversely, when significant differences in pressure fluctuations occur, the stability value drops rapidly, effectively revealing potential anomalies or emergencies. This mechanism overcomes the shortcomings of traditional simple differential or fixed threshold methods, such as sensitivity to noise and coarse judgment. It makes the regional situational stability assessment more robust, nuanced, and dynamically adaptable, providing a reliable reference for full-domain security pressure prediction, video response frequency calculation, and automatic resource scheduling, thereby improving the system's real-time performance, accuracy, and robustness.

[0012] Furthermore, the historical time window is set. for: , The size of the first preset time window. For normalization function, For the first Standard deviation of the overall security stress data at any given time and within its historical initial time window. This is the floor function. This is the reference time window size.

[0013] The aforementioned technical solution adaptively adjusts the size of historically set time windows based on the fluctuation level of overall security pressure data. This allows for the use of smaller time windows to improve response speed when pressure changes are stable, and larger time windows to enhance the stability of statistical results when pressure fluctuations are significant. This adaptive time window mechanism avoids the lag or oversensitivity issues associated with fixed time windows, making pressure sequence analysis more flexible and reliable. This improves the accuracy and dynamic adaptability of overall security situation assessment, providing a more robust data foundation for subsequent video response frequency prediction and automatic resource scheduling.

[0014] Furthermore, the size of the reference time window is specifically determined by: obtaining the first... The mode of all global security pressure data under the second preset time window corresponding to the time will be far from the first time. The time corresponding to the mode furthest from the time and the first time The difference in time is used as the reference time window size.

[0015] The aforementioned technical solution determines a reference time window based on typical values ​​of pressure data over a longer time range. The system can adaptively capture the periodic or recurring patterns of pressure sequences, thus balancing data representativeness and time span when calculating historical statistical characteristics. When pressure exhibits regular fluctuations, the reference time window can cover key change intervals, improving the stability and reliability of statistical results. Conversely, when pressure changes are irregular or experience sudden fluctuations, this mechanism can still provide a reasonable time scale through mode positioning, avoiding excessive influence of short-term anomalies on the analysis results, thereby enhancing the accuracy and robustness of overall security pressure prediction and video response frequency scheduling.

[0016] Furthermore, the video response frequency at the next moment for: , The mode of all global security pressure data within the historical time window corresponding to the current moment. The mode is the video response frequency. This provides data on the overall security pressure at the next moment.

[0017] The aforementioned technical solution uses typical values ​​of historical pressure data at the current moment as a reference benchmark and maps the corresponding video response frequency (VRF) to the pressure at the next moment. This establishes a stable correlation between pressure changes and historical patterns, making the prediction of VRF more reasonable and smooth. This method balances the representativeness of historical data with the real-time changes in current pressure, effectively avoiding the bias caused by direct linear extrapolation based on instantaneous pressure. This improves the accuracy and dynamic adaptability of VRF prediction, providing a reliable basis for the automatic scheduling of security resources and response to abnormal events.

[0018] Further, the acquisition of the overall security pressure data sequence is specifically as follows: the video frame difference, audio decibels and crowd density data of the overall security zone at each time point are acquired through the Internet of Things gateway, and normalized after digital signal conversion. The average value after normalization is calculated at each time point, and the average value is used as the security pressure data of the overall security zone at the corresponding time point to obtain the overall security pressure data sequence.

[0019] Furthermore, the preset hyperparameter is 0.1.

[0020] Furthermore, the size of the first preset time window is 20.

[0021] Furthermore, the size of the second preset time window is 100.

[0022] The beneficial effects of this invention are as follows: This invention generates a pressure sequence from multi-source security data across the entire defense zone, and uses an improved exponential smoothing algorithm combined with regional situational stability to adaptively predict the pressure at the next moment. The predicted pressure is then mapped to video response frequency for automatic scheduling, achieving dynamic optimization of camera acquisition parameters, analysis tasks, and bandwidth allocation. The adaptive smoothing coefficient and historical time window adjustment enable rapid response to significant pressure fluctuations and smoothness during stable periods. The video response frequency calculated based on typical pressure values ​​improves the sensitivity and accuracy of responses to abnormal events, thereby enhancing the real-time performance, intelligence, and reliability of the entire security resource scheduling, while effectively improving the overall system operating efficiency and security protection level. Attached Figure Description

[0023] Figure 1 is a flowchart schematically illustrating a dynamic scheduling method for global security resources according to an embodiment of the present invention. Detailed Implementation

[0024] Example of a method for dynamic scheduling of security resources across the entire domain.

[0025] As shown in Figure 1, the flowchart of the dynamic scheduling method for full-domain security resources according to an embodiment of the present invention includes the following steps: S1: Obtain the full-domain security pressure data sequence at the current moment and under its historical set time window.

[0026] In one embodiment, obtaining the overall security pressure data sequence at the current moment and within its historical set time window involves: acquiring video frame difference, audio decibels, and crowd density data of the entire security zone at each moment through an IoT gateway, and performing normalization processing after digital signal conversion. The average value after normalization is calculated at each moment, and the average value is used as the security pressure data of the entire security zone at the corresponding time node to obtain the overall security pressure data sequence.

[0027] By digitizing and normalizing multi-source sensing data collected across the entire security zone at various times, and using the comprehensive average of various data as the quantification result of security pressure at the corresponding time, the real-time changes in the overall security situation can be accurately reflected on a unified scale. This not only avoids evaluation biases caused by inconsistencies in the dimensions of different physical quantities, but also improves the stability and comparability of security pressure representation. This allows subsequent risk assessment, dynamic scheduling, and anomaly detection to be based on continuous, objective, and processable data sequences, thereby significantly improving the real-time performance, accuracy, and intelligence level of the overall security system.

[0028] In one embodiment, the historical setting time window for: , The size of the first preset time window. For normalization function, For the first Standard deviation of the overall security stress data at any given time and within its historical initial time window. This is the floor function. This is for reference time window size. The first preset time window size is 20, but it can be set according to the actual situation.

[0029] The reference time window size is specifically: obtaining the first... The mode of all global security pressure data under the second preset time window corresponding to the time will be far from the first time. The time corresponding to the mode furthest from the time and the first time The difference in time is used as the reference time window size. The second preset time window is 100, but it can also be set according to the actual situation.

[0030] By adaptively adjusting the time window size based on the fluctuation characteristics of comprehensive security pressure data and determining the reference time scale by combining the distribution characteristics of pressure data over a larger time range, an appropriate data statistical range can be dynamically matched according to the stability of the security situation at different stages. When pressure changes are stable, the time window automatically shrinks to improve response speed; when pressure fluctuations are large, the time window automatically expands to enhance the stability of statistical results. This avoids the lag or oversensitivity problems caused by fixed time windows, making pressure assessment more flexible, reliable, and able to more realistically reflect the changes in the comprehensive security status over time, thereby significantly improving the accuracy and robustness of subsequent analysis and decision-making.

[0031] S2: The global security pressure data sequence is smoothed using an improved exponential smoothing algorithm to obtain the global security pressure data at the next moment.

[0032] In one embodiment, the improved exponential smoothing algorithm includes a smoothing coefficient, the smoothing coefficient... for: , As the initial value, , The first Time, Number The stability of the regional situation at any given time.

[0033] By dynamically adjusting the smoothing coefficient based on changes in situational stability at adjacent time points, the exponential smoothing process automatically adjusts its sensitivity to new data according to the steady-state of the region. When the situation becomes unstable, the smoothing coefficient automatically increases, thereby enhancing the responsiveness to the latest pressure changes; conversely, when the situation remains stable, the smoothing coefficient automatically decreases to strengthen the smoothing effect and suppress the influence of occasional noise. This avoids the problem of insufficient adaptability of traditional fixed smoothing coefficients in different scenarios, enabling the processing of pressure sequences to possess both rapid responsiveness and high stability, thus significantly improving the accuracy and robustness of overall situational analysis.

[0034] The stability of the region for: , For the natural constant An exponential function with base 0. For the first Standard deviation of the overall security pressure data under the given time window and its historical settings. For the first Standard deviation of the overall security pressure data under the given time window and its historical settings. For the first The maximum value among the standard deviations of the overall security stress data for all times within a given time window and its historical settings. For the first The size of the historical time window corresponding to each moment. This is a preset hyperparameter. The preset hyperparameter is 0.1, but it can also be set according to the actual situation.

[0035] By constructing an exponential regional situational stability model based on the differences in pressure fluctuations at different time points, this model can reflect changes in regional security situation over time in a continuous, smooth, and sensitive manner. The closer historical pressure fluctuations are to current pressure characteristics, the higher the stability value, indicating a consistent regional state. Conversely, as the difference between the two increases, the stability value decreases rapidly, highlighting potential anomalies or sudden changes. This approach avoids the problems of noise sensitivity and coarse judgment found in traditional simple difference or fixed threshold methods, making stability assessment more robust, nuanced, and dynamically adaptable, providing a more reliable basis for subsequent risk identification, early warning triggering, and resource scheduling.

[0036] S3: Obtain the video response frequency for the next moment based on the overall security pressure data, and perform automatic scheduling based on the video response frequency.

[0037] In one embodiment, the video response frequency at the next moment for: , The mode of all global security pressure data within the historical time window corresponding to the current moment. The mode is the video response frequency. This provides data on the overall security pressure at the next moment.

[0038] By using the mode of the overall security pressure as a reference benchmark and proportionally mapping the pressure value at the next moment using its corresponding video response frequency, the system can establish a stable correlation between pressure fluctuations and historical typical situations, thereby achieving a more reasonable inference of future video response frequencies. This avoids the bias caused by directly extrapolating linearly based on instantaneous pressure, allowing the prediction results to simultaneously reflect the representativeness of historical patterns and the real-time nature of current pressure changes. This improves the accuracy and dynamic adaptability of video response frequency prediction, providing a more reliable basis for subsequent resource scheduling and risk warning.

[0039] Automatic scheduling is performed based on the video response frequency. Specifically, the video acquisition parameters of the target defense zone are automatically adjusted according to the degree and trend of changes in the video response frequency. Through this scheduling mechanism, monitoring and analysis capabilities can be improved in a timely manner when the video is dynamically enhanced, and resource consumption can be reduced when the video changes are stable. This enables adaptive allocation and dynamic optimization of security resources, improving the real-time performance, effectiveness, and resource utilization efficiency of the overall security response.

[0040] This invention generates a pressure sequence from multi-source security data across the entire defense zone. It then uses an improved exponential smoothing algorithm combined with regional situational stability to adaptively predict the pressure at the next moment. The predicted pressure is then mapped to video response frequency for automatic scheduling, achieving dynamic optimization of camera parameters, analysis tasks, and bandwidth allocation. The smoothing coefficient automatically adjusts with changes in situational stability, and the historical time window adaptively adjusts, ensuring rapid response during periods of significant fluctuation and stability during periods of calm. Furthermore, the video response frequency is calculated using typical pressure values, improving the sensitivity to anomaly detection. This effectively enhances the accuracy and stability of overall security pressure assessment, achieving real-time, intelligent, and efficient resource scheduling management.

[0041] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise explicitly specified.

[0042] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A method for dynamic scheduling of security resources across the entire domain, characterized in that, include: The system acquires the global security pressure data sequence at the current moment and within its historical set time window. An improved exponential smoothing algorithm is then used to smooth the global security pressure data sequence to obtain the global security pressure data for the next moment. Based on the global security pressure data, the video response frequency for the next moment is obtained, and automatic scheduling is performed based on the video response frequency. The improved exponential smoothing algorithm includes a smoothing coefficient, which is positively correlated with the initial value and the regional situation stability at the previous moment, and negatively correlated with the regional situation stability at the current moment.

2. The method for dynamic scheduling of all-domain security resources according to claim 1, characterized in that, The smoothness coefficient for: ， As the initial value, 、 The first Time, Number The stability of the regional situation at any given time.

3. The method for dynamic scheduling of all-domain security resources according to claim 1, characterized in that, The stability of the region for: ， For the natural constant An exponential function with base 0. For the first Standard deviation of the overall security pressure data under the specified time window and its historical settings. For the first Standard deviation of the overall security pressure data under the specified time window and its historical settings. For the first The maximum value among the standard deviations of the overall security stress data for all times within a given time window and its historical settings. For the first The size of the historical time window corresponding to each moment. These are the preset hyperparameters.

4. The method for dynamic scheduling of all-domain security resources according to claim 1, characterized in that, The historical time window for: ， The size of the first preset time window. For normalization function, For the first Standard deviation of the overall security stress data at any given time and within its historical initial time window. This is the floor function. This is the reference time window size.

5. The method for dynamic scheduling of all-domain security resources according to claim 4, characterized in that, The reference time window size is specifically: obtaining the first... The mode of all global security pressure data under the second preset time window corresponding to the time will be far from the first time. The time corresponding to the mode furthest from the time and the first time The difference in time is used as the reference time window size.

6. The method for dynamic scheduling of all-domain security resources according to claim 1, characterized in that, The video response frequency at the next moment for: ， The mode of all global security pressure data within the historical time window corresponding to the current moment. The mode is the video response frequency. This provides data on the overall security pressure at the next moment.

7. The method for dynamic scheduling of all-domain security resources according to claim 1, characterized in that, The process of obtaining the overall security pressure data sequence is as follows: the video frame difference, audio decibels, and crowd density data of the entire security zone at each time point are obtained through an IoT gateway, and then normalized after digital signal conversion. The average value of the normalized data is calculated at each time point, and the average value is used as the security pressure data of the entire security zone at the corresponding time point to obtain the overall security pressure data sequence.

8. The method for dynamic scheduling of all-domain security resources according to claim 3, characterized in that, The preset hyperparameter is 0.

1.

9. The method for dynamic scheduling of all-domain security resources according to claim 4, characterized in that, The size of the first preset time window is 20.

10. The method for dynamic scheduling of all-domain security resources according to claim 5, characterized in that, The second preset time window size is 100.

Images