Multi-terminal security management scheduling system

Abstract

The invention discloses a multi-terminal security management scheduling system, and relates to the technical field of security management. The system comprises a plurality of security monitoring terminals, a security management scheduling platform server and a scheduling execution terminal, and the server is used for extracting the visitor flow of a target parking lot from security monitoring data from the plurality of monitoring terminals in real time, generating visitor flow time sequence data, and then according to the visitor flow time sequence data, performing scheduling on the target parking lot according to the visitor flow time sequence data. The method comprises the following steps: predicting to obtain a people flow predicted value and a security demand index of a target place in a unit period in the current next day, then generating a security management scheduling strategy for the target place in the unit period in the current next day according to the demand index, and finally sending a scheduling command to an execution terminal according to the scheduling strategy. Therefore, a deployment scheme can be arranged no longer according to artificial experience, it is ensured that a dynamic arrangement result has a quantitative scientific basis, dead angles are effectively avoided, and actual security requirements can be dynamically met.

Description

Technical Field

[0001] The invention belongs to the technical field of security management, and in particular relates to a multi-terminal security management and dispatching system. Background Art

[0002] At present, for public places that require security presence, such as residential communities, large shopping malls, stadiums, and event sites, contracted security companies usually deploy a certain number of security personnel to conduct regular patrols and perform security work to avoid theft, robbery, physical conflicts, and natural disasters in the duty sites, so as to effectively prevent crimes and accidents and protect people's personal and property safety.

[0003] However, the existing security personnel deployment plans at the target duty sites (such as the specific number of security personnel deployed, etc.) are all arranged manually based on experience, lacking a quantitative scientific basis, and prone to blind spots, which may lead to a situation where the security personnel deployment plan does not match the actual security needs. For example, if there are too few security personnel deployed at a certain target duty site, it will cause incomplete patrol coverage and untimely hidden danger discovery and resolution. If there are too many security personnel deployed at a certain target duty site, it will cause unnecessary waste of human resources. In addition, the existing security personnel deployment plan is basically fixed within a day, lacking dynamic optimization technical means to dynamically adapt to actual security needs. Summary of the invention

[0004] The purpose of the present invention is to provide a multi-terminal security management and dispatching system to solve the problems that the existing security personnel deployment plan is arranged manually based on experience, resulting in a lack of quantitative scientific basis, prone to blind spots, and possible mismatch between the plan and actual security needs.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] In a first aspect, a multi-terminal security management and dispatching system is provided, comprising a plurality of security monitoring terminals, a security management and dispatching platform server and a dispatching execution terminal, wherein the security management and dispatching platform server is respectively communicatively connected to the plurality of security monitoring terminals and the dispatching execution terminal;

[0007] The security monitoring terminal is deployed at the target duty location, and is used to collect security monitoring data of the target duty location in real time, and transmit the security monitoring data to the security management and dispatching platform server in real time;

[0008] The security management and scheduling platform server is used to extract the flow of people at the target duty place in real time from the security monitoring data from the multiple security monitoring terminals, and generate time series data of the flow of people, and then predict the predicted value of the flow of people at the target duty place in the current next daily unit period according to the flow of people time series data, and then calculate the security demand index of the target duty place in the current next daily unit period according to the predicted flow of people and the preset security weight coefficient of the target duty place, and then generate a security management scheduling strategy for the target duty place in the current next daily unit period according to the security demand index and the preset security management scheduling strategy generation rule, and finally send a scheduling command to the scheduling execution terminal according to the security management scheduling strategy, wherein the flow of people time series data contains multiple daily unit periods that are consecutive in time sequence and the historical flow of people in each daily unit period among the multiple daily unit periods;

[0009] The dispatch execution terminal is used to execute the received dispatch command so as to go to or evacuate the target duty place.

[0010] Based on the above invention content, a new security management solution is provided for dynamically generating and executing security management scheduling strategies based on dynamic prediction results of human flow, namely, comprising multiple security monitoring terminals, a security management scheduling platform server and a scheduling execution terminal, wherein the server is used to extract the human flow of a target duty place in real time from the security monitoring data from multiple monitoring terminals, and generate human flow time series data, and then predict the human flow prediction value and security demand index of the target place in the unit period of the current next day according to the human flow time series data, and then generate a security management scheduling strategy for the unit period of the current next day for the target place according to the demand index, and finally send a scheduling command to the execution terminal according to the scheduling strategy, so that the execution terminal goes to or evacuates the target place, so that the security personnel deployment plan is no longer arranged based on manual experience, ensuring that the dynamic arrangement results have both quantitative scientific basis, effectively avoiding blind spots, and dynamically adapting to actual security needs, which is convenient for practical application and promotion.

[0011] In one possible design, the multiple security monitoring terminals include a traffic monitoring radar, an aerial photography drone, a surveillance camera, a law enforcement recorder held by a security personnel, a smartphone held by a security personnel, a tablet computer held by a security personnel and / or AR smart glasses worn on the head of a security personnel.

[0012] In one possible design, the dispatch execution terminal includes an aerial photography drone, a law enforcement recorder held by security personnel and used to instruct the holder to go to or evacuate the target duty location by displaying the received dispatch command, a smart phone held by security personnel and used to instruct the holder to go to or evacuate the target duty location by displaying the received dispatch command, a tablet computer held by security personnel and used to instruct the holder to go to or evacuate the target duty location by displaying the received dispatch command, and / or AR smart glasses worn on the head of the security personnel and used to instruct the wearer to go to or evacuate the target duty location by displaying the received dispatch command.

[0013] In a possible design, extracting the flow of people at the target duty place in real time from the security monitoring data from the multiple security monitoring terminals includes:

[0014] For each of the plurality of security monitoring terminals, extract the target duty place and the corresponding original human flow in real time from the corresponding security monitoring data, and generate corresponding original human flow time series data;

[0015] For each security monitoring terminal, according to the preset multiple unified collection moments in the unit time period, the collection moments in the corresponding original human flow time series data are shifted as a whole, so that each shifted collection moment corresponds to each moment in the multiple unified collection moments in the unit time period, and for each shifted collection moment, the corresponding aligned human flow time series data is calculated according to the following formula:

[0016]

[0017] In the formula, represents the total number of moments in the multiple unified collection moments within the unit time period, Indicates less than or equal to A positive integer, Indicates The aligned flow of people corresponding to the collected time after translation, Indicates the The time difference between the shifted acquisition moment and the first moment among the multiple unified acquisition moments in the unit time period, and Respectively represent the original flow of people in the corresponding time series data and One-to-one correspondence between adjacent acquisition moments The coefficients of the local cubic polynomial used to implement cubic spline interpolation are obtained by solving the original flow of people using the least squares method. represents a positive integer greater than or equal to 4, The adjacent recording time refers to the previous time closest to the nth post-translation acquisition time. A collection moment, represents a positive integer;

[0018] For each post-translation collection moment, the target on-duty location and the corresponding flow of people are calculated according to the following formula:

[0019]

[0020] Indicates the target station location and is related to the first The flow of people corresponding to the collection time after the translation, M represents the total number of the multiple security monitoring terminals, m represents a positive integer less than or equal to M, η m represents the preset weight coefficient of the mth security monitoring terminal among the multiple security monitoring terminals, and there is Indicates the mth security monitoring terminal and the The aligned passenger flow corresponding to the collection time after the translation.

[0021] In a possible design, based on the time series data of the flow of people, the predicted value of the flow of people at the target duty place in the unit period of the current next day is predicted, including:

[0022] Dividing the pedestrian flow time series data into a plurality of pedestrian flow time series data corresponding one to one to a plurality of time windows that are sequentially continuous in time series;

[0023] For the last time window among the multiple time windows, at least one set of time series data of human flow corresponding to at least one corresponding previous time window is used as a corresponding training data set, and a set of time series data of human flow corresponding to the corresponding window is used as a corresponding verification data set;

[0024] Based on the Prophet model, a real-time dynamic prediction model for passenger flow is established using the following formula:

[0025]

[0026] Where t represents time, A(t) represents the flow of people at time t, g(t) represents the trend term at time t, s(t) represents the seasonal term at time t, h(t) represents the holiday term at time t, ε(t) represents the error noise term at time t, Q represents a positive integer, q represents a positive integer less than or equal to q, l q (t) represents the linear regression term corresponding to the qth linear influence factor among the q linear influence factors at time t;

[0027] The growth mode of the trend term is set to a linear growth mode, and natural weeks and natural days are added to make the seasonal term change periodically, and the error noise term is also set to obey a mathematical expectation of μ and a variance of σ 2 Normal distribution of

[0028] Adding an impact weight adjustment factor to the holiday item to obtain a first model parameter vector consisting of weight adjustment parameters of all holiday intervals in the holiday item;

[0029] Acquire time series data of all human traffic-related influencing factors, and then dynamically screen all human traffic-related influencing factors for human traffic linear influencing factors based on the time series data, so as to determine human traffic-related influencing factors that meet the screening conditions as the linear influencing factors, and obtain a second model parameter vector composed of weight adjustment parameters in linear regression items of all the linear influencing factors;

[0030] Using the training data set and the verification data set corresponding to the last time window, the real-time dynamic prediction model of pedestrian flow is trained and optimized based on the particle swarm algorithm to obtain a verified model parameter optimization search result, wherein the model parameter optimization search result includes the first model parameter vector, the second model parameter vector, the mathematical expectation and the variance optimization search result;

[0031] The model parameter optimization search result is input into the real-time dynamic prediction model for human flow, and then the real-time dynamic prediction model for human flow is used to predict the predicted value of human flow at the target duty place in a unit period within the current next day.

[0032] In a possible design, an impact weight adjustment factor is added to the holiday item to obtain a first model parameter vector consisting of weight adjustment parameters of all holiday intervals in the holiday item, including:

[0033] Assume that there are C holiday intervals in the holiday item: H1, H2, ..., H k ,…,H C , then the holiday term h(t) = λ×Z(t), where the holiday interval contains a natural day or at least two natural days consecutive in time sequence, C represents a positive integer, k represents a positive integer less than or equal to C, H k represents the kth holiday interval, λ=(λ1,λ2,…,λ k ,…,λ C ), d k Indicates the total number of days in the kth holiday interval, j indicates less than or equal to d k A positive integer, λk,j represents the weight adjustment factor of the jth day in the kth holiday interval and has ɑ k represents the first weight adjustment parameter of the kth holiday interval, β k represents the second weight adjustment parameter of the kth holiday interval, Indicates that if t belongs to the kth holiday interval H k , then the value is 1, otherwise the value is zero;

[0034] The first weight adjustment parameters and the second weight adjustment parameters of the C holiday intervals are summarized to obtain a first model parameter vector composed of the weight adjustment parameters of all holiday intervals in the holiday item.

[0035] In a possible design, dynamically screening the linear influencing factors of human traffic for all human traffic related influencing factors based on the time series data, so as to determine the human traffic related influencing factors meeting the screening conditions as the linear influencing factors, includes:

[0036] Randomly select historical pedestrian flow of multiple natural days to form a first sample S1, and select data of the multiple natural days from the corresponding time series data as a second sample S2 for a certain pedestrian flow-related influencing factor among all the pedestrian flow-related influencing factors;

[0037] Performing a normal distribution KS check on the first sample S1 to calculate a first check statistic p value p1, and also performing the normal distribution KS check on the second sample S2 to calculate a second check statistic p value p2;

[0038] If the first check statistic p value p1 and the second check statistic p value p2 are both greater than the first preset threshold, the correlation coefficient r between the first sample S1 and the sample S2 is calculated according to the following formula: q :

[0039]

[0040] In the formula, i represents a positive integer, x i represents the i-th sample value in the first sample S1, y i represents the i-th sample value in the second sample S2, represents the sample mean of the first sample S1, represents the sample mean of the second sample S2;

[0041] If the correlation coefficient r qIf the value is greater than a second preset threshold, it is determined that the certain human flow related influencing factor meets the screening condition, and the certain human flow related influencing factor is determined as the linear influencing factor.

[0042] In a possible design, the training data set and the verification data set corresponding to the last time window are used to perform model training and parameter optimization on the real-time dynamic prediction model of pedestrian flow based on the particle swarm algorithm to obtain a verified model parameter optimization search result, including the following steps S2701 to S2712:

[0043] S2701. Initialize the algorithm parameters of the particle swarm algorithm including the population size N, the maximum number of iterations T, the self-learning factor C1 and the global learning factor C2, and randomly generate the initial position of each particle in the particle population, and then execute step S2702, wherein the initial position of each particle is randomly generated based on the following formula:

[0044]

[0045] Wherein, i' represents a positive integer, d represents a positive integer less than or equal to D, D represents the number of dimensions of the model parameter vector to be optimized, and the model parameter vector to be optimized includes the first model parameter vector to be optimized, the second model parameter vector, the mathematical expectation and the variance, represents the initial position of the i'th particle in the particle population on the dth dimension of the model parameter vector to be optimized, u c,d represents the upper limit of the model parameter search space in the dth dimension, l c,d represents the lower limit of the model parameter search space in the d-th dimension, and rand(0,1) represents a pure decimal random generator function;

[0046] S2702. For each particle, the corresponding initial position is input into the real-time dynamic prediction model of human flow as the model parameter search result, and then the training data set corresponding to the last time window is used to train the real-time dynamic prediction model of human flow, and the corresponding fitness F is calculated, and then step S2703 is executed, wherein the fitness F is calculated according to the following formula:

[0047]

[0048] Where n represents the total number of samples in the training data set, i represents a positive integer, represents the prediction output item calculated by importing the model input item of the i-th sample data in the training data set into the real-time dynamic prediction model of pedestrian flow, y (i) represents the true output item of the i-th sample data;

[0049] S2703. Take the initial position of a particle with the minimum fitness as the initial global optimal position G ini , the position of the i'th particle in the iteration process and corresponding to the minimum fitness is the local optimal position P reached by the i'th particle i’ , and the position of the particle population in the iteration process and corresponding to the minimum fitness is recorded as the global optimal position G reached by the particle population, and the moving speed of the i'th particle is recorded as V i’ , and finally initialize the local optimal position P i’ =G=G ini , and initialize the moving speed V i’ is zero, and the particle movement inertia coefficient range is also set to [T min ,T max ], and initialize the period T of the periodic reduction of the inertia coefficient to a randomly selected value within the range of the particle movement inertia coefficient, and also initialize the particle movement inertia coefficient in the iteration process to a value greater than zero, and also initialize the current iteration number t' to 0, and then execute step S2704;

[0050] S2704. Update the particle movement inertia coefficient of each particle, and then execute step S2705, wherein the update formula of the particle movement inertia coefficient is as follows:

[0051]

[0052] In the formula, represents the particle movement inertia coefficient of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the particle movement inertia coefficient of the i'th particle in the dth dimension before the t'+1th iteration, t'%T represents the modulo operation result of the previous iteration number t' on the period T, and if If it is less than 0, a new value is randomly selected within the range of the particle's inertia coefficient as the new period T, and the calculation is recalculated. until Not less than 0;

[0053] S2705. Update the moving speed of each particle, and then execute step S2706, wherein the updating formula of the moving speed is as follows:

[0054]

[0055] In the formula, represents the moving speed of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the moving speed of the i'th particle in the dth dimension before the t'+1th iteration, γ1 represents the first random number uniformly distributed in (0,1), represents the local optimal position reached by the i'th particle in the dth dimension during the first t' iterations, represents the position of the i'th particle in the dth dimension before the t'+1th iteration, γ2 represents a second random number uniformly distributed in (0,1), represents the global optimal position reached by the particle population in the d-th dimension during the first t' iterations;

[0056] S2706. Update the position of each particle, and then execute step S2707, wherein the update formula of the position is as follows:

[0057] In the formula, represents the position of the i'th particle in the dth dimension obtained after the t'+1th iteration;

[0058] S2707. For each particle, the corresponding position is input into the real-time dynamic prediction model of human flow as the model parameter search result, and then the training data set corresponding to the last time window is applied again to train the real-time dynamic prediction model of human flow, and the corresponding fitness F is calculated again, and then step S2708 is executed;

[0059] S2708. Determine the local optimal position P i’ Is the corresponding fitness greater than the current fitness of the i'th particle? If so, the local optimal position P i’ Update to the current position of the i'th particle, and then execute step S2709, otherwise directly execute step S2709;

[0060] S2709. Determine whether the fitness corresponding to the global optimal position G is greater than the minimum value among the current fitness of each particle. If so, update the global optimal position G to the current position of any particle with the minimum value, and then execute step S2710. Otherwise, directly execute step S2710;

[0061] S2710. Increment the current number of iterations t' by 1, and determine whether the current number of iterations t' reaches the maximum number of iterations T. If so, execute step S2711, otherwise return to execute step S2704;

[0062] S2711. Input the global optimal position G as the model parameter search result into the real-time dynamic prediction model for human flow, and then apply the real-time dynamic prediction model for human flow and the verification data set corresponding to the last time window to predict human flow, and calculate the new fitness F new , then execute step S2712;

[0063] S2712. Judgement F new <ρ×F min Is it true? If so, the global optimal position G is used as the verified model parameter optimization search result. Otherwise, after adjusting the search space of the model parameter vector to be optimized, return to step S2701, where ρ represents the preset error verification coefficient and is greater than 1, F min represents the fitness corresponding to the global optimal position G, and the model parameter optimization search result includes the optimization search results of the first model parameter vector, the second model parameter vector, the mathematical expectation and the variance.

[0064] In a possible design, the security demand index of the target duty place in the unit period of the current next day is calculated according to the predicted value of the flow of people and the preset security weight coefficient of the target duty place, including:

[0065] The result of multiplying the predicted value of the flow of people by the preset security weight coefficient of the target duty place is used as the security demand index of the target duty place in the unit period within the current next day.

[0066] In a possible design, when the dispatch execution terminal is a law enforcement recorder held by a security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, a smart phone held by a security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, a tablet held by a security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, or an AR smart glasses worn on the head of a security personnel and used to instruct the wearer to go to or evacuate the target duty place by displaying the received dispatch command, a security management dispatch strategy is generated for the target duty place in the unit period within the current next day according to the security demand index and the preset security management dispatch strategy generation rule, including:

[0067] According to the security demand index, the number of security personnel N required for the target duty location in the next day unit period is calculated according to the following formula:

[0068]

[0069] Where N totle represents the number of security personnel that can be provided in the unit period of the current next day, P represents the security demand index, ΣP represents the sum of the security demand indexes of all the target duty places in the unit period of the current next day, and Round() represents the rounding function;

[0070] According to the difference between the number of security personnel required by the target duty place in the unit period of the current day and N, a security management scheduling strategy in the unit period of the next current day is generated for the target duty place.

[0071] Beneficial effects of the above scheme:

[0072] (1) The present invention creatively provides a new security management solution for dynamically generating and executing security management scheduling strategies based on dynamic prediction results of human flow, namely, comprising multiple security monitoring terminals, a security management scheduling platform server and a scheduling execution terminal, wherein the server is used to extract the human flow of a target duty place from the security monitoring data from multiple monitoring terminals in real time, and generate human flow time series data, and then predict the human flow prediction value and security demand index of the target place in the current next day unit period based on the human flow time series data, and then generate a security management scheduling strategy for the target place in the current next day unit period based on the demand index, and finally send a scheduling command to the execution terminal based on the scheduling strategy, so that the execution terminal goes to or evacuates the target place, so that the security personnel deployment plan is no longer arranged based on manual experience, ensuring that the dynamic arrangement results have both quantitative scientific basis, effectively avoiding blind spots, and can dynamically adapt to actual security needs, which is convenient for practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0074] Figure 1 A schematic diagram of the structure of a multi-terminal security management and dispatching system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0075] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structures of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other embodiments can be obtained based on these embodiments without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0076] It should be understood that although the terms first and second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another object. For example, a first object can be referred to as a second object, and similarly, a second object can be referred to as a first object without departing from the scope of the exemplary embodiments of the present invention.

[0077] It should be understood that the term "and / or" that may appear in this document is merely a description of the association relationship between associated objects, indicating that there may be three relationships. For example, A and / or B can indicate three situations: A exists alone, B exists alone, or A and B exist at the same time. For another example, A, B and / or C can indicate the existence of any one of A, B and C or any combination of them. The term " / and" that may appear in this document describes another type of association object relationship, indicating that there may be two relationships. For example, A / and B can indicate two situations: A exists alone or A and B exist at the same time. In addition, the character " / " that may appear in this document generally indicates that the previous and next associated objects are in an "or" relationship.

[0078] Example

[0079] like Figure 1 As shown, the multi-terminal security management and dispatching system provided in this embodiment includes but is not limited to multiple security monitoring terminals, a security management and dispatching platform server, and a dispatching execution terminal, etc., wherein the security management and dispatching platform server is respectively connected to the multiple security monitoring terminals and the dispatching execution terminal. The specific method of the security management and dispatching platform server communicating with each terminal can adopt an existing wired communication method or an existing wireless communication method, such as a 4G / 5G wireless communication method.

[0080] The security monitoring terminal is deployed at the target duty place, and is used to collect the security monitoring data of the target duty place in real time, and transmit the security monitoring data to the security management and dispatching platform server in real time. The target duty place can be, but is not limited to, a public place such as a residential community with security duty needs, a large shopping mall, a gymnasium, an event site, or a specific location in the aforementioned places (such as the entrance and exit of the event site, etc.). Specifically, the multiple security monitoring terminals include but are not limited to people flow monitoring radars, aerial photography drones, surveillance cameras, law enforcement recorders held by security personnel, smart phones held by security personnel, tablet computers held by security personnel and / or AR (Augmented Reality) smart glasses worn on the heads of security personnel, etc., wherein the people flow monitoring radar can collect radar data used to reflect the degree of human flow as the security monitoring data, which can be specifically implemented by existing millimeter wave radars or lidars; the aerial photography drones, the surveillance cameras, the law enforcement recorders, the smart phones, the tablet computers or the AR smart glasses can collect video data used to extract the degree of human flow as the security monitoring data, which can be specifically implemented by existing related electronic products.

[0081] The security management and scheduling platform server is used to extract the flow of people at the target duty place in real time from the security monitoring data from the multiple security monitoring terminals, and generate time series data of the flow of people, and then predict the predicted value of the flow of people at the target duty place in the next unit period of the day in the future based on the time series data of the flow of people, and then calculate the security demand index of the target duty place in the next unit period of the day in the future based on the predicted value of the flow of people and the preset security weight coefficient of the target duty place, and then generate a security management scheduling strategy for the target duty place in the next unit period of the day in the future based on the security demand index and the preset security management scheduling strategy generation rule, and finally send a scheduling command to the scheduling execution terminal according to the security management scheduling strategy, wherein the time series data of the flow of people contains multiple unit time periods of the day that are consecutive in time sequence and the historical flow of people in each unit time period of the day in the multiple unit time periods.

[0082] The security management and scheduling platform server is used to extract the flow of people at the target duty place in real time from the security monitoring data from the multiple security monitoring terminals, and generate time series data of the flow of people, and then predict the predicted value of the flow of people at the target duty place in the current next daily unit period based on the flow of people time series data, and then calculate the security demand index of the target duty place in the current next daily unit period based on the predicted flow of people and the preset security weight coefficient of the target duty place, and then generate a security management scheduling strategy for the target duty place in the current next daily unit period according to the security demand index and the preset security management scheduling strategy generation rule, and finally send a scheduling command to the scheduling execution terminal according to the security management scheduling strategy, wherein the flow of people time series data contains multiple daily unit time periods that are consecutive in time sequence and the historical flow of people in each daily unit time period among the multiple daily unit time periods. The above-mentioned specific method of extracting human flow in real time from security monitoring data is an existing technical means, such as conventionally extracting human flow from radar data used to reflect the degree of human flow, and using the existing YOLO (You Only Look Once, currently the latest version has been developed to V5, and is also widely used in the industry. Its basic principle is: first, the input image is divided into 7×7 grids, and 2 borders are predicted for each grid. Then, according to the threshold, target windows with relatively low probability are removed, and finally, the redundant windows are removed by border merging to obtain the detection result) target detection algorithm to identify the human body marker frame from the video frame image of the video data, and then obtain the human flow based on the statistical results of the human body marker frame. The intra-day unit time period can be specifically determined based on the longest reaction time of the security personnel to or from the target duty place, which can be, but not limited to, 1 hour. All the recorded historical human flow can be divided into sub-intervals in the time dimension: starting from the early morning of each day, each day is divided into 24 time periods with a period of 1 hour, and the average value of the human flow in each time period, the median value of the human flow or the maximum value of the human flow, etc., are counted, and then the human flow statistics of multiple days are combined into the human flow time series data. Specifically, according to the predicted human flow value and the preset security weight coefficient of the target duty place, the security demand index of the target duty place in the unit period of the current next day is calculated, including but not limited to: the multiplication result of the predicted human flow value and the preset security weight coefficient of the target duty place is used as the security demand index of the target duty place in the unit period of the current next day. Specifically, the security management scheduling strategy generation rules specifically include but are not limited to: the greater the security demand index, the greater the number of security personnel deployed, and vice versa. The security management scheduling strategy generated in this way will specifically increase or decrease the number of security personnel deployed.

[0083] The above-mentioned sending of a dispatch command to the dispatch execution terminal according to the security management dispatch strategy specifically includes but is not limited to the following: (1) when it is necessary to increase the number of deployed security personnel, determine whether the security management dispatch strategy of other nearby target duty places in the unit period of the current next day is to reduce the number of deployed security personnel; if so, send a dispatch command for instructing the holder or wearer to go to the target duty place to the dispatch execution terminal held or worn by some security personnel currently located in the other nearby target duty places (for example, the first security personnel, the first two security personnel, or the first few security personnel, etc., selected in order from the smallest to the largest distance from the current position of the personnel to the target duty place); (2) when it is necessary to reduce the number of deployed security personnel, send a dispatch command for instructing the holder or wearer to go to the target duty place to the dispatch execution terminal held or worn by some security personnel currently located in the target duty place (for example, the first security personnel, the first two security personnel, or the first few security personnel, etc., selected in order from the smallest to the largest distance from the current position of the personnel to the target duty place); The scheduling execution terminal held or worn by the first security personnel, the first two security personnel, or the first few security personnel, etc.) sends a scheduling command for instructing the holder or wearer to evacuate the target duty place; (3) when the number of security personnel deployment that needs to be increased exceeds the first preset number threshold, it is determined whether the security management scheduling strategy of other nearby target duty places in the current next day unit period is to reduce the number of security personnel deployment and whether the number of security personnel deployment that needs to be reduced exceeds the second preset number threshold. If both are true, a scheduling command for instructing the drone to go to the target duty place is sent to the aerial photography drone currently deployed in the other nearby target duty places; (4) when the number of security personnel deployment that needs to be reduced exceeds the second preset number threshold, a scheduling command for instructing the drone to evacuate the target duty place is sent to the aerial photography drone currently deployed in the target duty place; and so on. Based on the sending mechanism of the aforementioned scheduling command, it can be ensured that the security personnel deployment plan of this embodiment has a certain quantitative scientific basis to avoid blind spots.

[0084] The dispatch execution terminal is used to execute the received dispatch command so as to go to or evacuate the target duty place. Specifically, the dispatch execution terminal includes but is not limited to an aerial photography drone, a law enforcement recorder held by a security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, a smart phone held by a security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, a tablet computer held by a security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, and / or AR smart glasses worn on the head of the security personnel and used to instruct the wearer to go to or evacuate the target duty place by displaying the received dispatch command, etc. In this way, the dispatch execution terminal can go to or evacuate the target duty place in an active manner, such as when the dispatch execution terminal is the aerial photography drone, or it can go to or evacuate the target duty place in a passive manner, such as when the dispatch execution terminal is the law enforcement recorder, the smart phone, the tablet computer or the AR smart glasses, etc.

[0085] Therefore, based on the aforementioned multi-terminal security management and dispatching system, a new security management solution is provided for dynamically generating and executing security management dispatching strategies based on dynamic prediction results of human flow, namely, it includes multiple security monitoring terminals, a security management dispatching platform server and a dispatching execution terminal, wherein the server is used to extract the human flow of the target duty place in real time from the security monitoring data from multiple monitoring terminals, and generate human flow time series data, and then predict the human flow prediction value and security demand index of the target place in the current unit period of the next day according to the human flow time series data, and then generate a security management dispatching strategy for the target place in the current unit period of the next day according to the demand index, and finally send a dispatching command to the execution terminal according to the dispatching strategy, so that the execution terminal goes to or evacuates the target place, so that the security personnel deployment plan is no longer arranged based on manual experience, ensuring that the dynamic arrangement results have both quantitative scientific basis, effectively avoiding blind spots, and dynamically adapting to actual security needs, which is convenient for practical application and promotion.

[0086] Preferably, considering that adjacent collection moments of different security monitoring terminals may have the characteristic of equal time intervals (because security monitoring data are collected periodically, for example, every 50 milliseconds), in order to better maintain the consistency of the collection timing of the pedestrian flow time series data obtained based on the multiple security monitoring terminals and comprehensively improve the accuracy of the pedestrian flow extraction results, the pedestrian flow of the target duty place is extracted in real time from the security monitoring data from the multiple security monitoring terminals, including but not limited to the following steps S11 to S13.

[0087] S11. For each security monitoring terminal among the plurality of security monitoring terminals, extract the target duty place and the corresponding original human flow in real time from the corresponding security monitoring data, and generate corresponding original human flow time series data.

[0088] In step S11, the original human flow time series data contains a plurality of collection moments that are sequentially timed and the historical human flow at each of the plurality of collection moments, wherein the aforementioned collection moment is specifically the moment when the corresponding security monitoring data is collected.

[0089] S12. For each of the security monitoring terminals, according to the preset multiple unified collection moments in the unit time period, the collection moments in the corresponding original human flow time series data are shifted as a whole, so that each shifted collection moment corresponds to each moment in the multiple unified collection moments in the unit time period, and for each shifted collection moment, the corresponding aligned human flow time series data is calculated according to the following formula:

[0090]

[0091] In the formula, represents the total number of moments in the multiple unified collection moments within the unit time period, Indicates less than or equal to A positive integer, Indicates The aligned flow of people corresponding to the collected time after translation, Indicates the The time difference between the shifted acquisition moment and the first moment among the multiple unified acquisition moments in the unit time period, and Respectively represent the original flow of people in the corresponding time series data and One-to-one correspondence between adjacent acquisition moments The coefficients of the local cubic polynomial used to implement cubic spline interpolation are obtained by solving the original flow of people using the least squares method. represents a positive integer greater than or equal to 4, The adjacent recording time refers to the previous time closest to the nth post-translation acquisition time. A collection moment, Represents a positive integer.

[0092] In the step S12, for example, if the multiple unified collection moments in the preset unit time period are respectively the end moment of 50 milliseconds, the end moment of 100 milliseconds, ..., the end moment of 950 milliseconds and the end moment of 1000 milliseconds in the unit time period (that is, the unit time period is 1 second), and the collection moments in the original human flow time series data obtained based on a security monitoring terminal are respectively 10:00:00:60 milliseconds, 10:00:00:110 milliseconds, ..., 10:00:00:960 milliseconds and 10:00:00:1 0:01:10 milliseconds, then after the collection moments in the original pedestrian flow time series data are shifted forward by 10 milliseconds as a whole, the collection moments after the shift can be made to be 10:00:00:50 milliseconds, 10:00:00:100 milliseconds, ..., 10:00:00:950 milliseconds and 10:00:01:0 millisecond, respectively, and they correspond one by one and are aligned with the end moment of the 50th millisecond, the end moment of the 100th millisecond, ..., the end moment of the 950th millisecond and the end moment of the 1000th millisecond in the unit time period.

[0093] S13. For each post-translation collection moment, the target station location and the corresponding flow of people are calculated according to the following formula:

[0094]

[0095] Indicates the target station location and is related to the first The flow of people corresponding to the collection time after the translation, M represents the total number of the multiple security monitoring terminals, m represents a positive integer less than or equal to M, η m represents the preset weight coefficient of the mth security monitoring terminal among the multiple security monitoring terminals, and there is Indicates the mth security monitoring terminal and the The aligned passenger flow corresponding to the collection time after the translation.

[0096] Preferably, in order to improve the accuracy of the pedestrian flow prediction value, the pedestrian flow prediction value of the target duty place in the unit period of the current next day is predicted based on the pedestrian flow time series data, including but not limited to the following steps S21 to S28.

[0097] S21. Divide the pedestrian flow time series data into a plurality of pedestrian flow time series data corresponding one to one to a plurality of time windows that are consecutive in time series.

[0098] In the step S21, the time window may specifically be, but is not limited to, a natural day, a plurality of consecutive natural days, or a week.

[0099] S22. For the last time window among the multiple time windows, at least one set of time series data on human traffic corresponding to at least one previous time window is used as a corresponding training data set, and a set of time series data on human traffic corresponding to the corresponding window is used as a corresponding verification data set.

[0100] In step S22, the last time window may specifically be but not limited to today or this week; the at least one previous time window may specifically be but not limited to yesterday and the day before yesterday, etc., or the at least one previous time window may specifically be but not limited to last week and the week before last, etc.

[0101] S23. Based on the Prophet model, a real-time dynamic prediction model for passenger flow is established using the following formula:

[0102]

[0103] Where t represents time, A(t) represents the flow of people at time t, g(t) represents the trend term at time t, s(t) represents the seasonal term at time t, h(t) represents the holiday term at time t, ε(t) represents the error noise term at time t, Q represents a positive integer, q represents a positive integer less than or equal to Q, l q (t) represents the linear regression term at time t and corresponding to the qth linear influence factor among the Q linear influence factors.

[0104] In step S23, the Prophet model is an open source time series prediction tool developed by the Artificial Intelligence Research Institute. It is based on an additive model and can process data with time and seasonal effects. It can be applied to time series prediction problems in various fields, including but not limited to financial market predictions, sales data analysis, and disease incidence monitoring. It provides an easy-to-use and highly flexible framework that allows even non-professional data analysts to effectively perform time series predictions. Therefore, the real-time dynamic prediction model of passenger flow can be obtained based on the regular modification of the Prophet model.

[0105] S24. Set the growth mode of the trend term to a linear growth mode, add natural weeks and natural days to make the seasonal term change periodically, and also set the error noise term to obey a mathematical expectation of μ and a variance of σ 2 The normal distribution of .

[0106] In the step S24, the specific configuration technical details of the trend term, the seasonal term and the error noise term can be conventionally derived by referring to the corresponding term configuration means of the existing Prophet model, and will not be repeated here.

[0107] S25. Add an impact weight adjustment factor to the holiday item to obtain a first model parameter vector consisting of weight adjustment parameters of all holiday intervals in the holiday item.

[0108] In the step S25, specifically, an influencing weight adjustment factor is added to the holiday item to obtain a first model parameter vector composed of weight adjustment parameters of all holiday intervals in the holiday item, including but not limited to the following steps S251 to S252.

[0109] S251. Assume that there are C holiday intervals in the holiday item: H1, H2, ..., H k ,…,H C , then the holiday term h(t) = λ×Z(t), where the holiday interval contains a natural day or at least two natural days consecutive in time sequence, C represents a positive integer, k represents a positive integer less than or equal to C, H k represents the kth holiday interval, λ=(λ1,λ2,…,λ k ,…,λ C ), d k Indicates the total number of days in the kth holiday interval, j indicates less than or equal to d k A positive integer, λ k,j represents the weight adjustment factor of the jth day in the kth holiday interval and has ɑ k represents the first weight adjustment parameter of the kth holiday interval, β k represents the second weight adjustment parameter of the kth holiday interval, Indicates that if t belongs to the kth holiday interval H k , then the value is 1, otherwise the value is zero.

[0110] S252. Summarize the first weight adjustment parameters and the second weight adjustment parameters of the C holiday intervals to obtain a first model parameter vector composed of the weight adjustment parameters of all holiday intervals in the holiday item.

[0111] S26. Obtaining time series data of all human traffic-related influencing factors, and then dynamically screening all human traffic-related influencing factors based on the time series data as human traffic linear influencing factors, so as to determine the human traffic-related influencing factors that meet the screening conditions as the linear influencing factors, and obtaining a second model parameter vector consisting of weight adjustment parameters in the linear regression items of all the linear influencing factors;

[0112] In the step S26, the human flow related influencing factors include but are not limited to any possible related factors such as the daily maximum temperature, the daily maximum humidity and / or the regional population density, and their time series data can be conventionally obtained. Specifically, based on the time series data, the human flow linear influencing factors are dynamically screened for the human flow related influencing factors, so that the human flow related influencing factors that meet the screening conditions are determined as the linear influencing factors, including but not limited to the following steps S261 to S264.

[0113] S261. Randomly select historical pedestrian flow of multiple natural days to form a first sample S1, and for a certain pedestrian flow-related influencing factor among all the pedestrian flow-related influencing factors, select the data of the multiple natural days from the corresponding time series data as a second sample S2.

[0114] In the step S261, the plurality of natural days may be greater than 50 days, for example.

[0115] S262. Perform a normal distribution KS check on the first sample S1 to calculate a first check statistic p value p1, and also perform the normal distribution KS check on the second sample S2 to calculate a second check statistic p value p2.

[0116] S263. If the first check statistic p value p1 and the second check statistic p value p2 are both greater than the first preset threshold, the correlation coefficient r between the first sample S1 and the sample S2 is calculated according to the following formula: q :

[0117]

[0118] In the formula, i represents a positive integer, x i represents the i-th sample value in the first sample S1, y i represents the i-th sample value in the second sample S2, represents the sample mean of the first sample S1, represents the sample mean of the second sample S2.

[0119] In the step S263, the first preset threshold may be 0.05, for example.

[0120] S264. If the correlation coefficient r q If the value is greater than a second preset threshold, it is determined that the certain human flow related influencing factor meets the screening condition, and the certain human flow related influencing factor is determined as the linear influencing factor.

[0121] In the step S264, the second preset threshold may be 0.3, for example.

[0122] In the step S26, for the certain flow-related influencing factor determined as the linear influencing factor, the corresponding linear regression term l can be specifically recorded as q (t) = r q ×θ q ×M q (t), where θ q Represents the corresponding weight adjustment parameter, M q (t) represents the corresponding numerical value, so that the second model parameter vector composed of weight adjustment parameters in the linear regression terms of all the linear influencing factors can be obtained.

[0123] S27. Apply the training data set and the verification data set corresponding to the last time window, and perform model training and parameter optimization on the real-time dynamic prediction model of pedestrian flow based on the particle swarm algorithm to obtain a verified model parameter optimization search result, wherein the model parameter optimization search result includes the optimization search results of the first model parameter vector, the second model parameter vector, the mathematical expectation and the variance.

[0124] In step S27, the particle swarm optimization (PSO) is a heuristic optimization algorithm that simulates the information sharing and collaboration mechanism in the foraging behavior of bird flocks, and finds the optimal solution to the problem through interaction between individuals. In order to balance the local optimum and global optimum capabilities in parameter optimization, improve the efficiency of optimizing parameter search while ensuring overall reliability, preferably, a training data set and a verification data set corresponding to the last time window are used, and the real-time dynamic prediction model of human flow is trained and optimized based on the particle swarm algorithm to obtain a verified model parameter optimization search result, including but not limited to the following steps S2701 to S2712.

[0125] S2701. Initialize the algorithm parameters of the particle swarm algorithm including the population size N, the maximum number of iterations T, the self-learning factor C1 and the global learning factor C2, and randomly generate the initial position of each particle in the particle population, and then execute step S2702, wherein the initial position of each particle is randomly generated based on the following formula:

[0126]

[0127] Wherein, i' represents a positive integer, d represents a positive integer less than or equal to D, D represents the number of dimensions of the model parameter vector to be optimized, and the model parameter vector to be optimized includes the first model parameter vector to be optimized, the second model parameter vector, the mathematical expectation and the variance, represents the initial position of the i'th particle in the particle population on the dth dimension of the model parameter vector to be optimized, u c,d represents the upper limit of the model parameter search space in the dth dimension, l c,d represents the lower limit of the model parameter search space in the d-th dimension, and rand(0,1) represents a pure decimal random generation function.

[0128] S2702. For each particle, the corresponding initial position is input into the real-time dynamic prediction model of human flow as the model parameter search result, and then the training data set corresponding to the last time window is used to train the real-time dynamic prediction model of human flow, and the corresponding fitness F is calculated, and then step S2703 is executed, wherein the fitness F is calculated according to the following formula:

[0129]

[0130] Where n represents the total number of samples in the training data set, i represents a positive integer, represents the prediction output item calculated by importing the model input item of the i-th sample data in the training data set into the real-time dynamic prediction model of pedestrian flow, y (i) Represents the true output item of the i-th sample data.

[0131] In step S2702, the specific process of the model training is the existing technology.

[0132] S2703. Take the initial position of a particle with the minimum fitness as the initial global optimal position G ini , the position of the i'th particle in the iteration process and corresponding to the minimum fitness is the local optimal position P reached by the i'th particle i’ , and the position of the particle population in the iteration process and corresponding to the minimum fitness is recorded as the global optimal position G reached by the particle population, and the moving speed of the i'th particle is recorded as V i’ , and finally initialize the local optimal position P i’ =G=G ini , and initialize the moving speed V i’ is zero, and the particle movement inertia coefficient range is also set to [T min ,Y max ], and initialize the period T of the periodic reduction of the inertia coefficient to a randomly selected value within the range of the particle movement inertia coefficient, and also initialize the particle movement inertia coefficient during the iteration process to a value greater than zero, and also initialize the current iteration number t' to 0, and then execute step S2704.

[0133] S2704. Update the particle movement inertia coefficient of each particle, and then execute step S2705, wherein the update formula of the particle movement inertia coefficient is as follows:

[0134]

[0135] In the formula, represents the particle movement inertia coefficient of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the particle movement inertia coefficient of the i'th particle in the dth dimension before the t'+1th iteration, t'%T represents the modulo operation result of the previous iteration number t' on the period T, and if If it is less than 0, a new value is randomly selected within the range of the particle's inertia coefficient as the new period T, and the calculation is recalculated. until Not less than 0.

[0136] S2705. Update the moving speed of each particle, and then execute step S2706, wherein the updating formula of the moving speed is as follows:

[0137]

[0138] In the formula, represents the moving speed of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the moving speed of the i'th particle in the dth dimension before the t'+1th iteration, γ1 represents the first random number uniformly distributed in (0,1), represents the local optimal position reached by the i'th particle in the dth dimension during the first t' iterations, represents the position of the i'th particle in the dth dimension before the t'+1th iteration, γ2 represents a second random number uniformly distributed in (0,1), represents the global optimal position reached by the particle population in the d-th dimension during the first t' iterations.

[0139] S2706. Update the position of each particle, and then execute step S2707, wherein the update formula of the position is as follows:

[0140] In the formula, represents the position of the i'th particle in the dth dimension obtained after the t'+1th iteration.

[0141] S2707. For each particle, the corresponding position is input into the real-time dynamic prediction model for pedestrian flow as the model parameter search result, and then the training data set corresponding to the last time window is applied again to train the real-time dynamic prediction model for pedestrian flow, and the corresponding fitness F is calculated again, and then step S2708 is executed.

[0142] S2708. Determine the local optimal position P i’ Is the corresponding fitness greater than the current fitness of the i'th particle? If so, the local optimal position P i’ Update to the current position of the i'th particle, and then execute step S2709; otherwise, directly execute step S2709.

[0143] S2709. Determine whether the fitness corresponding to the global optimal position G is greater than the minimum value among the current fitness of each particle. If so, update the global optimal position G to the current position of any particle with the minimum value, and then execute step S2710; otherwise, directly execute step S2710.

[0144] S2710. Increment the current number of iterations t' by 1, and determine whether the current number of iterations t' reaches the maximum number of iterations T. If so, execute step S2711; otherwise, return to execute step S2704.

[0145] S2711. Input the global optimal position G as the model parameter search result into the real-time dynamic prediction model for human flow, and then apply the real-time dynamic prediction model for human flow and the verification data set corresponding to the last time window to predict human flow, and calculate the new fitness F new , then execute step S2712.

[0146] S2712. Judgement F new <ρ×F min Is it true? If so, the global optimal position G is used as the verified model parameter optimization search result. Otherwise, after adjusting the search space of the model parameter vector to be optimized, return to step S2701, where ρ represents the preset error verification coefficient and is greater than 1, F min represents the fitness corresponding to the global optimal position G, and the model parameter optimization search result includes the optimization search results of the first model parameter vector, the second model parameter vector, the mathematical expectation and the variance.

[0147] In step S27, the particle swarm algorithm may be replaced by a Newton optimization algorithm, a genetic optimization algorithm, a Grey Wolf Algorithm, a whale optimization algorithm or a tuna swarm optimization algorithm. Considering that different optimization algorithms have different advantages and disadvantages, in order to combine the performance of the aforementioned optimization algorithms to achieve the purpose of making the best use of their strengths and avoiding their weaknesses, this embodiment further preferably uses a training data set and a verification data set corresponding to the last time window, and performs model training and parameter optimization on the real-time dynamic prediction model of human flow according to the following steps S27001 to S27018 to obtain a verified model parameter optimization search result.

[0148] S27001. Initialize the optimization algorithm parameters including the number of search individuals I in the search population, the maximum number of iterations T, the first learning factor α1, the second learning factor α2, the third learning factor α3, the fourth learning factor α4, the fifth learning factor α5 and the sixth learning factor α6, and randomly generate the initial position of each search individual in the search population, and then execute step S27002, wherein the initial position of each search individual is randomly generated based on the following formula:

[0149]

[0150] Wherein, i′ represents a positive integer less than or equal to I, d′ represents a positive integer less than or equal to D′, D′ represents the number of dimensions of the parameter vector to be optimized, and the model parameter vector to be optimized includes the first model parameter vector to be optimized, the second model parameter vector, the mathematical expectation and the variance, represents the component of the initial position of the i′th search individual in the search population on the d′th dimension of the parameter vector to be optimized, u c,d′ represents the upper limit of the parameter search space in the d′th dimension, l c,d′ represents the lower limit of the parameter search space in the d′th dimension, and rand(0,1) represents a pure decimal random generation function.

[0151] S27002. For each search individual, the corresponding initial position is input as the model parameter search result into the real-time dynamic prediction model for human flow, and then the training data set corresponding to the last time window is used to train the real-time dynamic prediction model for human flow, and the corresponding fitness is calculated, and then step S27003 is executed, wherein the fitness F is calculated according to the following formula:

[0152]

[0153] Where n represents the total number of samples in the training data set, i represents a positive integer, represents the prediction output item calculated by importing the model input item of the i-th sample data in the training data set into the real-time dynamic prediction model of pedestrian flow, y (i) Represents the true output item of the i-th sample data.

[0154] In step S27002, the specific process of the model training is the existing technology.

[0155] S27003. Take the initial position of a search individual with the minimum fitness as the initial global optimal position x best , and also initialize the current number of iterations t′=0, and then execute step S27004.

[0156] S27004. Calculate the current average position of the search population according to the current position of each search individual, and determine the corresponding first new position for each search individual, and then execute step S27005, wherein the first new position of each search individual is determined according to the following formula:

[0157] x new1,i′,d′ =x best,d′ +α1×rand(0,1)×(x mean,d′ -x i′,d′ )

[0158] In the formula, x new1,i′,d′ represents the component of the first new position of the i′th search individual in the d′th dimension, x best,d′ Denotes the global optimal position x best The component in the d′th dimension, x mean,d′ represents the component of the current average position of the search population in the d′th dimension, x i′,d′ Represents the component of the current position of the i′th search individual in the d′th dimension.

[0159] S27005. For each of the search individuals, the corresponding first new position is input into the real-time dynamic prediction model for pedestrian flow as the model parameter search result, and then the training data set corresponding to the last time window is applied again to train the real-time dynamic prediction model for pedestrian flow, and the corresponding and new fitness is calculated again, and then step S27006 is executed.

[0160] S27006. Determine whether the fitness corresponding to the current position of the i′th search individual is greater than the current fitness of the i′th search individual. If so, update the current position of the i′th search individual to the first new position of the i′th search individual, and then execute step S27007; otherwise, directly execute step S27007.

[0161] S27007. Determine the global optimal position x best Is the corresponding fitness greater than the minimum value among the current fitness of each search individual? If so, the global optimal position x best Update to the current position of any search individual with the minimum value, and then execute step S27008, otherwise directly execute step S27008.

[0162] S27008. Update the calculation according to the current position of each search individual to obtain the current average position of the search population, and determine the corresponding second new position for each search individual, and then execute step S27009, wherein the second new position of each search individual is determined according to the following formula:

[0163]

[0164] In the formula, x new2,i′,d′ represents the component of the second new position of the i′th search individual in the d′th dimension, β(i′) represents the first intermediate variable corresponding to the i′th search individual, δ(i′) represents the second intermediate variable corresponding to the i′th search individual, i″ represents a positive integer less than or equal to I, mod() represents a remainder function, x i″,d′ represents the component of the current position of the i′-th search individual in the search population on the d′-th dimension, βr(i′) represents the third intermediate variable corresponding to the i′-th search individual, δr(i′) represents the fourth intermediate variable corresponding to the i′-th search individual, max() represents the maximum value function, θ(i′) represents the fifth intermediate variable corresponding to the i′-th search individual, r(i′) represents the sixth intermediate variable corresponding to the i′-th search individual, and π represents 180 degrees.

[0165] S27009. For each of the search individuals, the corresponding second new position is input into the real-time dynamic prediction model for pedestrian flow as the model parameter search result, and then the training data set corresponding to the last time window is applied again to train the real-time dynamic prediction model for pedestrian flow, and the corresponding and new fitness is calculated again, and then step S27010 is executed.

[0166] S27010. Determine whether the fitness corresponding to the current position of the i′th search individual is greater than the current fitness of the i′th search individual. If so, update the current position of the i′th search individual to the second new position of the i′th search individual, and then execute step S27011; otherwise, directly execute step S27011.

[0167] S27011. Determine the global optimal position x best Is the corresponding fitness greater than the minimum value among the current fitness of each search individual? If so, the global optimal position x best Update to the current position of any search individual with the minimum value, and then execute step S27012, otherwise directly execute step S27012.

[0168] S27012. The current average position of the search population is calculated and updated according to the current position of each search individual, and the corresponding third new position is determined for each search individual, and then step S27013 is executed, wherein the third new position of each search individual is determined according to the following formula:

[0169]

[0170] In the formula, x new3,i′,d′ represents the component of the third new position of the i′th search individual in the d′th dimension, represents the seventh intermediate variable corresponding to the i′th search individual, represents the eighth intermediate variable corresponding to the i′th search individual, represents the ninth intermediate variable corresponding to the i′th search individual, represents the tenth intermediate variable corresponding to the i′th search individual, Represents the eleventh intermediate variable corresponding to the i′th search individual.

[0171] S27013. For each of the search individuals, the corresponding third new position is input into the real-time dynamic prediction model for pedestrian flow as the model parameter search result, and then the training data set corresponding to the last time window is applied again to train the real-time dynamic prediction model for pedestrian flow, and the corresponding and new fitness is calculated again, and then step S27014 is executed.

[0172] S27014. Determine whether the fitness corresponding to the current position of the i′th search individual is greater than the current fitness of the i′th search individual. If so, update the current position of the i′th search individual to the third new position of the i′th search individual, and then execute step S27015; otherwise, directly execute step S27015.

[0173] S27015. Determine the global optimal position x best Is the corresponding fitness greater than the minimum value among the current fitness of each search individual? If so, the global optimal position x best Update to the current position of any search individual with the minimum value, and then execute step S27016, otherwise directly execute step S27016.

[0174] S27016. Increment the current number of iterations t′ by 1, and determine whether the current number of iterations t′ reaches the maximum number of iterations T. If so, execute step S27017, otherwise return to execute step S27004.

[0175] S27017. The global optimal position x best The real-time dynamic prediction model for human flow is input as the model parameter search result, and then the real-time dynamic prediction model for human flow and the validation data set corresponding to the last time window are applied to perform human flow prediction, and a new fitness F is calculated. new , then execute step S27018.

[0176] S27018. Judgement F new <ρ×F min Is it true? If so, then the global optimal position x best As the verified model parameter optimization search result, otherwise after adjusting the search space of the model parameter vector to be optimized, return to step S27001, where ρ represents the preset error verification coefficient and is greater than 1, F min represents the fitness corresponding to the global optimal position G, and the model parameter optimization search result includes the optimization search results of the first model parameter vector, the second model parameter vector, the mathematical expectation and the variance.

[0177] Based on the aforementioned steps S27001 to S27018, since each search individual performs three fitness calculations with the help of spirals and group centers in each iteration process, it can be regarded as a fusion of the particle swarm optimization algorithm and the whale optimization algorithm. Therefore, under all the same conditions, its optimization performance will be better than other optimization algorithms, which is conducive to quickly and accurately obtaining verified model parameter optimization search results.

[0178] S28. Input the model parameter optimization search result into the real-time dynamic prediction model for human flow, and then use the real-time dynamic prediction model for human flow to predict the predicted value of human flow at the target duty place in the next day unit period.

[0179] In the step S28, for example, if the current intraday unit time period is from 12:00 to 13:00, then the current next intraday unit time period is from 13:00 to 14:00.

[0180] Based on the above steps S21 to S28, the pedestrian flow prediction value can have the characteristics of high prediction real-time performance, high prediction accuracy and stronger model generalization ability, which is further convenient for practical application and promotion.

[0181] Specifically, when the dispatch execution terminal is a law enforcement recorder held by security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, a smart phone held by security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, a tablet computer held by security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, or AR smart glasses worn on the head of the security personnel and used to instruct the wearer to go to or evacuate the target duty place by displaying the received dispatch command, a security management dispatch strategy for the unit period within the current next day is generated for the target duty place according to the security demand index and the preset security management dispatch strategy generation rules, including but not limited to S31~S32.

[0182] S31. According to the security demand index, the number of security personnel N required for the target duty location in the next day unit period is calculated according to the following formula:

[0183]

[0184] Where N totle represents the number of security personnel that can be provided in the unit period within the current next day, P represents the security demand index, ΣP represents the sum of the security demand indexes of all the target duty places in the unit period within the current next day, and Round() represents the rounding function.

[0185] In step S31, the above formula reflects the security management scheduling strategy generation rule: the greater the security demand index, the greater the number of security personnel deployed, and vice versa.

[0186] S32. Generate a security management scheduling strategy for the target duty location in the unit period of the next current day according to the difference between the number of security personnel required by the target duty location in the unit period of the current day and N.

[0187] In summary, the multi-terminal security management and dispatching system provided by this embodiment has the following technical effects:

[0188] (1) This embodiment provides a new security management solution for dynamically generating and executing security management scheduling strategies based on dynamic prediction results of human flow, namely, comprising multiple security monitoring terminals, a security management scheduling platform server and a scheduling execution terminal, wherein the server is used to extract the human flow of a target duty place in real time from the security monitoring data from multiple monitoring terminals, and generate human flow time series data, and then predict the human flow prediction value and security demand index of the target place in the current next day unit period based on the human flow time series data, and then generate a security management scheduling strategy for the target place in the current next day unit period based on the demand index, and finally send a scheduling command to the execution terminal based on the scheduling strategy, so that the execution terminal goes to or evacuates the target place. In this way, the security personnel deployment plan is no longer arranged based on manual experience, ensuring that the dynamic arrangement result has both a quantitative scientific basis, effectively avoiding blind spots, and dynamically adapting to actual security needs, which is convenient for practical application and promotion.

[0189] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A multi-terminal security management and dispatching system, characterized in that: It includes multiple security monitoring terminals, a security management and dispatching platform server and a dispatching execution terminal, wherein the security management and dispatching platform server is respectively connected to the multiple security monitoring terminals and the dispatching execution terminal in communication; The security monitoring terminal is deployed at the target duty location, and is used to collect security monitoring data of the target duty location in real time, and transmit the security monitoring data to the security management and dispatching platform server in real time; The security management and scheduling platform server is used to extract the flow of people at the target duty place in real time from the security monitoring data from the multiple security monitoring terminals, and generate time series data of the flow of people, and then predict the predicted value of the flow of people at the target duty place in the current next daily unit period according to the flow of people time series data, and then calculate the security demand index of the target duty place in the current next daily unit period according to the predicted flow of people and the preset security weight coefficient of the target duty place, and then generate a security management scheduling strategy for the target duty place in the current next daily unit period according to the security demand index and the preset security management scheduling strategy generation rule, and finally send a scheduling command to the scheduling execution terminal according to the security management scheduling strategy, wherein the flow of people time series data contains multiple daily unit periods that are consecutive in time sequence and the historical flow of people in each daily unit period among the multiple daily unit periods; The dispatch execution terminal is used to execute the received dispatch command so as to go to or evacuate the target duty place.

2. The multi-terminal security management and dispatching system according to claim 1, characterized in that: The multiple security monitoring terminals include pedestrian flow monitoring radars, aerial photography drones, surveillance cameras, law enforcement recorders held by security personnel, smart phones held by security personnel, tablet computers held by security personnel and / or AR smart glasses worn on the heads of security personnel.

3. The multi-terminal security management and dispatching system according to claim 1, characterized in that: The dispatch execution terminal includes an aerial photography drone, a law enforcement recorder held by security personnel and used to instruct the holder to go to or evacuate the target duty location by displaying the received dispatch command, a smart phone held by security personnel and used to instruct the holder to go to or evacuate the target duty location by displaying the received dispatch command, a tablet computer held by security personnel and used to instruct the holder to go to or evacuate the target duty location by displaying the received dispatch command, and / or AR smart glasses worn on the head of the security personnel and used to instruct the wearer to go to or evacuate the target duty location by displaying the received dispatch command.

4. The multi-terminal security management and dispatching system according to claim 1, characterized in that: Extracting the flow of people at the target duty place in real time from the security monitoring data from the multiple security monitoring terminals includes: For each of the plurality of security monitoring terminals, extract the target duty place and the corresponding original human flow in real time from the corresponding security monitoring data, and generate corresponding original human flow time series data; For each security monitoring terminal, according to the preset multiple unified collection moments in the unit time period, the collection moments in the corresponding original human flow time series data are shifted as a whole, so that each shifted collection moment corresponds to each moment in the multiple unified collection moments in the unit time period, and for each shifted collection moment, the corresponding aligned human flow time series data is calculated according to the following formula: In the formula, represents the total number of moments in the multiple unified collection moments within the unit time period, Indicates less than or equal to A positive integer, Indicates The aligned flow of people corresponding to the collected time after translation, Indicates the The time difference between the shifted acquisition moment and the first moment among the multiple unified acquisition moments in the unit time period, and Respectively represent the original flow of people in the corresponding time series data and One-to-one correspondence between adjacent acquisition moments The coefficients of the local cubic polynomial used to implement cubic spline interpolation are obtained by solving the original flow of people using the least squares method. represents a positive integer greater than or equal to 4, The adjacent recording time refers to the previous time closest to the nth post-translation acquisition time. A collection moment, represents a positive integer; For each post-translation collection moment, the target on-duty location and the corresponding flow of people are calculated according to the following formula: Indicates the target station location and is related to the first The flow of people corresponding to the collection time after the translation, M represents the total number of the multiple security monitoring terminals, m represents a positive integer less than or equal to M, η m represents the preset weight coefficient of the mth security monitoring terminal among the multiple security monitoring terminals, and there is Indicates the mth security monitoring terminal and the The aligned passenger flow corresponding to the collection time after the translation.

5. The multi-terminal security management and dispatching system according to claim 1, characterized in that: According to the time series data of the flow of people, the predicted value of the flow of people at the target duty place in the unit period of the current next day is predicted, including: Dividing the pedestrian flow time series data into a plurality of pedestrian flow time series data corresponding one to one to a plurality of time windows that are sequentially continuous in time series; For the last time window among the multiple time windows, at least one set of time series data of human flow corresponding to at least one corresponding previous time window is used as a corresponding training data set, and a set of time series data of human flow corresponding to the corresponding window is used as a corresponding verification data set; Based on the Prophet model, a real-time dynamic prediction model for passenger flow is established using the following formula: Where t represents time, A(t) represents the flow of people at time t, g(t) represents the trend term at time t, s(t) represents the seasonal term at time t, h(t) represents the holiday term at time t, ε(t) represents the error noise term at time t, Q represents a positive integer, q represents a positive integer less than or equal to Q, l q (t) represents the linear regression term corresponding to the qth linear influence factor among the Q linear influence factors at time t; The growth mode of the trend term is set to a linear growth mode, and natural weeks and natural days are added to make the seasonal term change periodically, and the error noise term is also set to obey a mathematical expectation of μ and a variance of σ 2 Normal distribution of Adding an impact weight adjustment factor to the holiday item to obtain a first model parameter vector consisting of weight adjustment parameters of all holiday intervals in the holiday item; Acquire time series data of all human traffic-related influencing factors, and then dynamically screen all human traffic-related influencing factors for human traffic linear influencing factors based on the time series data, so as to determine human traffic-related influencing factors that meet the screening conditions as the linear influencing factors, and obtain a second model parameter vector composed of weight adjustment parameters in linear regression items of all the linear influencing factors; Using the training data set and the verification data set corresponding to the last time window, the real-time dynamic prediction model of pedestrian flow is trained and optimized based on the particle swarm algorithm to obtain a verified model parameter optimization search result, wherein the model parameter optimization search result includes the first model parameter vector, the second model parameter vector, the mathematical expectation and the variance optimization search result; The model parameter optimization search result is input into the real-time dynamic prediction model for human flow, and then the real-time dynamic prediction model for human flow is used to predict the predicted value of human flow at the target duty place in a unit period within the current next day.

6. The multi-terminal security management and dispatching system according to claim 5, characterized in that: Adding an impact weight adjustment factor to the holiday item to obtain a first model parameter vector consisting of weight adjustment parameters of all holiday intervals in the holiday item includes: Assume that there are C holiday intervals in the holiday item: H1, H2, ..., H k ,…,H C , then the holiday term h(t) = λ×Z(t), where the holiday interval contains a natural day or at least two natural days consecutive in time sequence, C represents a positive integer, k represents a positive integer less than or equal to C, H k represents the kth holiday interval, λ=(λ1,λ2,…,λ k ,…,λ C ), d k Indicates the total number of days in the kth holiday interval, j indicates less than or equal to d k A positive integer, λ k,j represents the weight adjustment factor of the jth day in the kth holiday interval and has ɑ k represents the first weight adjustment parameter of the kth holiday interval, β k represents the second weight adjustment parameter of the kth holiday interval, Indicates that if t belongs to the kth holiday interval H k , then the value is 1, otherwise the value is zero; The first weight adjustment parameters and the second weight adjustment parameters of the C holiday intervals are summarized to obtain a first model parameter vector composed of the weight adjustment parameters of all holiday intervals in the holiday item.

7. The multi-terminal security management and dispatching system according to claim 5, characterized in that: Dynamically screening the linear influencing factors of human traffic for all human traffic related influencing factors based on the time series data, so as to determine the human traffic related influencing factors meeting the screening conditions as the linear influencing factors, including: Randomly select historical pedestrian flow of multiple natural days to form a first sample S1, and select data of the multiple natural days from the corresponding time series data as a second sample S2 for a certain pedestrian flow-related influencing factor among all the pedestrian flow-related influencing factors; Performing a normal distribution KS check on the first sample S1 to calculate a first check statistic p value p1, and also performing the normal distribution KS check on the second sample S2 to calculate a second check statistic p value p2; If the first check statistic p value p1 and the second check statistic p value p2 are both greater than the first preset threshold, the correlation coefficient r between the first sample S1 and the sample S2 is calculated according to the following formula: q : In the formula, i represents a positive integer, x i represents the i-th sample value in the first sample S1, y i represents the i-th sample value in the second sample S2, represents the sample mean of the first sample S1, represents the sample mean of the second sample S2; If the correlation coefficient r q If the value is greater than a second preset threshold, it is determined that the certain human flow related influencing factor meets the screening condition, and the certain human flow related influencing factor is determined as the linear influencing factor.

8. The multi-terminal security management and dispatching system according to claim 5, characterized in that: Using the training data set and the verification data set corresponding to the last time window, the real-time dynamic prediction model of pedestrian flow is trained and optimized based on the particle swarm algorithm to obtain the verified model parameter optimization search result, including the following steps S2701 to S2712: S2701. Initialize the algorithm parameters of the particle swarm algorithm including the population size N, the maximum number of iterations T, the self-learning factor C1 and the global learning factor C2, and randomly generate the initial position of each particle in the particle population, and then execute step S2702, wherein the initial position of each particle is randomly generated based on the following formula: Wherein, i' represents a positive integer, d represents a positive integer less than or equal to D, D represents the number of dimensions of the model parameter vector to be optimized, and the model parameter vector to be optimized includes the first model parameter vector to be optimized, the second model parameter vector, the mathematical expectation and the variance, represents the initial position of the i'th particle in the particle population on the dth dimension of the model parameter vector to be optimized, u c,d represents the upper limit of the model parameter search space in the dth dimension, l c,d represents the lower limit of the model parameter search space in the d-th dimension, and rand(0,1) represents a pure decimal random generator function; S2702. For each particle, the corresponding initial position is input into the real-time dynamic prediction model of human flow as the model parameter search result, and then the training data set corresponding to the last time window is used to train the real-time dynamic prediction model of human flow, and the corresponding fitness F is calculated, and then step S2703 is executed, wherein the fitness F is calculated according to the following formula: Where n represents the total number of samples in the training data set, i represents a positive integer, represents the prediction output item calculated by importing the model input item of the i-th sample data in the training data set into the real-time dynamic prediction model of pedestrian flow, y (i) represents the true output item of the i-th sample data; S2703. Take the initial position of a particle with the minimum fitness as the initial global optimal position G ini , the position of the i'th particle in the iteration process and corresponding to the minimum fitness is the local optimal position P reached by the i'th particle i’ , and the position of the particle population in the iteration process and corresponding to the minimum fitness is recorded as the global optimal position G reached by the particle population, and the moving speed of the i'th particle is recorded as V i’ , and finally initialize the local optimal position P i’ =G=G ini , and initialize the moving speed V i’ is zero, and the particle movement inertia coefficient range is also set to [T min ,T max ], and initialize the period T of the periodic reduction of the inertia coefficient to a randomly selected value within the range of the particle movement inertia coefficient, and also initialize the particle movement inertia coefficient in the iteration process to a value greater than zero, and also initialize the current iteration number t' to 0, and then execute step S2704; S2704. Update the particle movement inertia coefficient of each particle, and then execute step S2705, wherein the update formula of the particle movement inertia coefficient is as follows: In the formula, represents the particle movement inertia coefficient of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the particle movement inertia coefficient of the i'th particle in the dth dimension before the t'+1th iteration, t'%T represents the modulo operation result of the previous iteration number t' on the period T, and if If it is less than 0, a new value is randomly selected within the range of the particle's inertia coefficient as the new period T, and the calculation is recalculated. until Not less than 0; S2705. Update the moving speed of each particle, and then execute step S2706, wherein the updating formula of the moving speed is as follows: In the formula, represents the moving speed of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the moving speed of the i'th particle in the dth dimension before the t'+1th iteration, γ1 represents the first random number uniformly distributed in (0,1), represents the local optimal position reached by the i'th particle in the dth dimension during the first t' iterations, represents the position of the i'th particle in the dth dimension before the t'+1th iteration, γ2 represents a second random number uniformly distributed in (0,1), represents the global optimal position reached by the particle population in the d-th dimension during the first t' iterations; S2706. Update the position of each particle, and then execute step S2707, wherein the update formula of the position is as follows: In the formula, represents the position of the i'th particle in the dth dimension obtained after the t'+1th iteration; S2707. For each particle, the corresponding position is input into the real-time dynamic prediction model of human flow as the model parameter search result, and then the training data set corresponding to the last time window is applied again to train the real-time dynamic prediction model of human flow, and the corresponding fitness F is calculated again, and then step S2708 is executed; S2708. Determine the local optimal position P i’ Is the corresponding fitness greater than the current fitness of the i'th particle? If so, the local optimal position P i’ Update to the current position of the i'th particle, and then execute step S2709, otherwise directly execute step S2709; S2709. Determine whether the fitness corresponding to the global optimal position G is greater than the minimum value among the current fitness of each particle. If so, update the global optimal position G to the current position of any particle with the minimum value, and then execute step S2710. Otherwise, directly execute step S2710; S2710. Increment the current number of iterations t' by 1, and determine whether the current number of iterations t' reaches the maximum number of iterations T. If so, execute step S2711, otherwise return to execute step S2704; S2711. Input the global optimal position G as the model parameter search result into the real-time dynamic prediction model for human flow, and then apply the real-time dynamic prediction model for human flow and the verification data set corresponding to the last time window to predict human flow, and calculate the new fitness F new , then execute step S2712; S2712. Judgement F new <ρ×F min Is it true? If so, the global optimal position G is used as the verified model parameter optimization search result. Otherwise, after adjusting the search space of the model parameter vector to be optimized, return to step S2701, where ρ represents the preset error verification coefficient and is greater than 1, F min represents the fitness corresponding to the global optimal position G, and the model parameter optimization search result includes the optimization search results of the first model parameter vector, the second model parameter vector, the mathematical expectation and the variance.

9. The multi-terminal security management and dispatching system according to claim 1, characterized in that: The security demand index of the target duty place in the unit period within the current next day is calculated according to the predicted value of the flow of people and the preset security weight coefficient of the target duty place, including: The result of multiplying the predicted value of the flow of people by the preset security weight coefficient of the target duty place is used as the security demand index of the target duty place in the unit period within the current next day.

10. The multi-terminal security management and dispatching system according to claim 1, characterized in that: When the dispatch execution terminal is a law enforcement recorder held by a security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, a smart phone held by a security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, a tablet held by a security personnel and used to instruct the holder to go to or evacuate the target duty place by displaying the received dispatch command, or an AR smart glasses worn on the head of a security personnel and used to instruct the wearer to go to or evacuate the target duty place by displaying the received dispatch command, a security management dispatch strategy is generated for the target duty place in the unit period within the current next day according to the security demand index and the preset security management dispatch strategy generation rule, including: According to the security demand index, the number of security personnel N required for the target duty location in the current next day unit period is calculated according to the following formula: Where N totle represents the number of security personnel that can be provided in the unit period of the current next day, P represents the security demand index, ∑P represents the sum of the security demand indexes of all the target duty places in the unit period of the current next day, and Round() represents the rounding function; According to the difference between the number of security personnel required by the target duty place in the unit period of the current day and N, a security management scheduling strategy in the unit period of the next current day is generated for the target duty place.

Images