A multi-line switching alarm method for intelligent traffic balancing control of radio and television
By calculating line stability and redefining the distribution distance in the broadcast and TV intelligent system, adaptively adjusting the shrinkage factor of the CURE clustering algorithm, the problem of poor multi-line switching effect in the broadcast and TV system is solved, and the traffic volume is improved The accuracy and intelligence of allocation.
Patent Information
- Application Number
- CN202410487009.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-22
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-04-22
AI Technical Summary
The existing clustering method based on representative points (CURE) cannot adaptively adjust the shrinkage factor in broadcast and television intelligent systems, resulting in poor multi-line switching effects and ineffective response to different density distribution conditions, affecting signal quality and user experience.
By obtaining the traffic data of different lines, calculating the line stability value and comprehensive stability at each moment, redefining the distribution distance between sample points, using the adaptive CURE clustering algorithm for clustering, and dynamically adjusting the shrinkage factor to improve clustering Class adaptability and accuracy.
It improves the accuracy and intelligence of multi-line switching, ensures the flexibility and accuracy of traffic allocation, enhances the clustering ability of traffic conditions at different pressure levels, and improves signal quality and user experience.
Smart Images

Figure CN118590682B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of line flow control, and in particular to a multi-line switching alarm method for intelligent flow balancing control of broadcasting and television. Background Art
[0002] In the broadcasting and television industry, intelligent traffic balancing control and multi-line switching are one of the key technical challenges. With the development of digital technology and the diversification of media content, traditional broadcasting and television networks are facing more and more traffic and signal processing requirements.
[0003] The broadcasting and television industry is facing challenges such as growing traffic demand, diversified content delivery methods, network planning and capacity management. It is particularly important to detect the operation status of the line in real time. Excessive traffic will directly affect the quality and continuity of the signal, and thus affect the user's viewing experience. The existing clustering method based on representative points (clustering using representative, CURE) classifies and detects the real-time data of different lines in the broadcasting and television intelligent system to judge the operation performance of the current line. However, the traditional CURE clustering algorithm uses a fixed shrinkage factor for clustering processing and cannot adapt the shrinkage factor to different density distribution conditions, which affects the adaptability and accuracy of the clustering algorithm, resulting in poor multi-line switching effect of broadcasting and television intelligent traffic. Summary of the invention
[0004] In order to solve the problem that the existing CURE clustering algorithm cannot adaptively determine the shrinkage factor when clustering the real-time data of different lines in the broadcasting and television intelligent system, thereby resulting in poor multi-line switching effect of the broadcasting and television intelligent traffic, the purpose of the present invention is to provide a multi-line switching alarm method for broadcasting and television intelligent traffic balance control, and the technical scheme adopted is as follows:
[0005] The present invention provides a multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television, the method comprising the following steps:
[0006] Obtain different types of traffic data for different lines in the radio and television intelligent system at each moment in the current time period and each sub-time period in each historical time period;
[0007] According to the fluctuation of each type of traffic data at each moment and the same type of traffic data in its neighborhood, the line stability value corresponding to each type of traffic data at each moment is obtained; according to the difference between each type of traffic data at each moment in each sub-time period in the current time period and the line stability value corresponding to the same type of traffic data in the corresponding sub-time period in the historical time period, and each type of traffic data at each moment in each sub-time period in each historical time period, the comprehensive stability of each type of traffic data at each moment in each sub-time period in the current time period is obtained;
[0008] Based on all kinds of traffic data at each moment in the current time period, sample points at each moment are obtained; according to the difference in comprehensive stability of each kind of traffic data at different sample points and the difference in similarity of different kinds of traffic data at different sample points and their neighborhoods, distribution distances between different sample points are obtained; based on the distribution distances, all sample points are clustered using the CURE clustering algorithm; in the clustering process, clusters with a number of sample points greater than a preset number are recorded as initial clusters; a contraction factor of each initial cluster is obtained based on the relative distance of sample points within each initial cluster; clustering is continued based on the contraction factor to obtain each target cluster;
[0009] Based on the number of sample points in each target cluster, it is determined whether to switch lines and issue an alarm.
[0010] Preferably, obtaining the line stability value corresponding to each type of flow data at each moment according to the fluctuation of each type of flow data at each moment and the same type of flow data in its neighborhood includes:
[0011] For the j-th type of traffic data at the i-th moment:
[0012] Perform curve fitting on the j-th flow data at the i-th moment and the j-th flow data at each moment in the neighborhood of the i-th moment to obtain the curve corresponding to the j-th flow data at the i-th moment;
[0013] According to the difference in data values between extreme points on the curve corresponding to the j-th type of flow data at the i-th moment, the line stability value corresponding to the j-th type of flow data at the i-th moment is obtained.
[0014] Preferably, the line stability value corresponding to the j-th flow data at the i-th moment is calculated using the following formula:
[0015]
[0016] Among them, Q i,jrepresents the line stability value corresponding to the j-th flow data at the ith moment, M represents the number of extreme points on the curve corresponding to the j-th flow data at the ith moment, Δq(m,m+2) represents the difference in data values between the m-th extreme point and the m+2-th extreme point on the curve corresponding to the j-th flow data at the ith moment, Δq(m+1,m+3) represents the difference in data values between the m+1-th extreme point and the m+3-th extreme point on the curve corresponding to the j-th flow data at the ith moment, σ(i,j) represents the standard deviation of the difference in data values between every two adjacent extreme points on the curve corresponding to the j-th flow data at the ith moment, norm() represents the normalization function, || represents the absolute value sign, λ1 is the preset first adjustment parameter, and λ1 is greater than 0.
[0017] Preferably, the comprehensive stability of each type of flow data at each moment in each sub-time period in the current time period is obtained according to the difference between each type of flow data at each moment in each sub-time period in the current time period and the line stability value corresponding to the same type of flow data in the corresponding sub-time period in the historical time period, and each type of flow data at each moment in each sub-time period in each historical time period, including:
[0018] For the j-th type of traffic data at the b-th moment in the k-th sub-time period of the current time period:
[0019] The average value of the line stability value corresponding to the j-th flow data at all moments in the k-th sub-time period in each historical time period is recorded as the first average value corresponding to the k-th sub-time period in each historical time period;
[0020] The absolute value of the difference between the b-th moment in the k-th sub-time period in each historical time period and the corresponding first average value is recorded as the first difference corresponding to the k-th sub-time period in each historical time period;
[0021] The absolute value of the difference between the line stability value corresponding to the j-th flow data at the b-th moment in the k-th sub-time period in the current time period and the line stability value corresponding to the j-th flow data at the b-th moment in the k-th sub-time period in each historical time period is recorded as the second difference corresponding to the b-th moment in the k-th sub-time period in each historical time period;
[0022] According to the first difference, the second difference and the line stability value corresponding to the bth moment in the kth sub-time period in the current time period, the comprehensive stability of the jth type of flow data at the bth moment in the kth sub-time period in the current time period is obtained.
[0023] Preferably, the following formula is used to calculate the comprehensive stability of the j-th type of flow data at the b-th moment in the k-th sub-time period in the current time period:
[0024]
[0025] Among them, W k,b,j represents the comprehensive stability of the jth type of traffic data at the bth moment in the kth sub-time period in the current time period, F represents the number of historical time periods, Q′ k,b,j Indicates the line stability value corresponding to the j-th flow data at the b-th time in the k-th sub-time period in the current time period, It represents the average value of the line stability value corresponding to the j-th flow data at all times in the k-th sub-time period in the f-th historical time period, Q″ f,k,b,j It represents the line stability value corresponding to the j-th flow data at the b-th moment in the k-th sub-time period in the f-th historical time period. exp() represents an exponential function with a natural constant as the base, and || represents the absolute value sign.
[0026] Preferably, the method of obtaining the distribution distances between different sample points according to the difference in comprehensive stability of each type of flow data at different sample points and the difference in similarity of different types of flow data at different sample points and their neighborhoods includes:
[0027] For the p-th sample point and the q-th sample point:
[0028] For the jth type of traffic data: the difference in comprehensive stability between the jth type of traffic data of the pth sample point and the qth sample point is recorded as the third difference corresponding to the jth type of traffic data of the pth sample point and the qth sample point; the jth type of traffic data of the pth sample point and the jth type of traffic data at all times in the neighborhood of the time corresponding to the pth sample point constitute the jth type of traffic data sequence corresponding to the pth sample point, and the jth type of traffic data of the qth sample point and the jth type of traffic data at all times in the neighborhood of the time corresponding to the qth sample point constitute the jth type of traffic data sequence corresponding to the qth sample point;
[0029] According to the third difference corresponding to each flow data of the p-th sample point and the q-th sample point, and the difference between the correlation between each two flow data sequences corresponding to the p-th sample point and the correlation between each two flow data sequences corresponding to the q-th sample point, the distribution distance between the p-th sample point and the q-th sample point is obtained.
[0030] Preferably, the method of obtaining the distribution distance between the p-th sample point and the q-th sample point according to the third difference corresponding to each flow data of the p-th sample point and the q-th sample point, and the difference between the correlation between each two flow data sequences corresponding to the p-th sample point and the correlation between each two flow data sequences corresponding to the q-th sample point, comprises:
[0031] Calculate the arithmetic square root of the sum of squares of the third differences corresponding to all types of traffic data of the p-th sample point and the q-th sample point;
[0032] According to the difference between the correlation between the arithmetic square root and each two traffic data sequences corresponding to the p-th sample point and the correlation between each two traffic data sequences corresponding to the q-th sample point, the distribution distance between the p-th sample point and the q-th sample point is obtained, and the difference between the arithmetic square root and the correlation between each two traffic data sequences is positively correlated with the distribution distance.
[0033] Preferably, obtaining a shrinkage factor of each initial cluster according to the relative distance of sample points in each initial cluster, and continuing clustering based on the shrinkage factor to obtain each target cluster, comprises:
[0034] For the cth initial cluster:
[0035] Calculate the cumulative sum of the distances between each sample point in the cth initial cluster and all other sample points in the cth initial cluster as the distance evaluation value of each sample point in the cth initial cluster; take the sample point corresponding to the minimum distance evaluation value as the target point;
[0036] Calculate the average distance between all sample points and the target point in the cth initial cluster; take a circular area with the target point as the center and the average distance as the radius as the reference area corresponding to the cth initial cluster;
[0037] A shrinkage factor of the c th initial cluster is obtained according to the difference between the distance between each sample point in the c th initial cluster and the target point and the average distance, the average distance, and the number of sample points in the reference area.
[0038] Preferably, the shrinkage factor of the cth initial cluster is calculated using the following formula:
[0039]
[0040] Among them, R c represents the shrinkage factor of the cth initial cluster, N represents the number of sample points in the cth initial cluster, represents the average distance between all sample points and the target point in the cth initial cluster, d c,n represents the distance between the nth sample point and the target point in the cth initial cluster, N′ represents the number of sample points in the reference area corresponding to the cth initial cluster, norm() represents the normalization function, and || takes the absolute value sign.
[0041] Preferably, the determining whether to perform line switching and alarm based on the number of sample points in each target cluster includes:
[0042] The number of sample points in the target cluster with the least number of sample points is recorded as the first number, and the number of sample points in the target cluster with the largest number of sample points is recorded as the second number;
[0043] If the ratio of the first number to the second number is greater than or equal to the preset proportion threshold, the line will not be switched and no alarm will be issued; if the ratio of the first number to the second number is less than the preset proportion threshold, the corresponding line will be adjusted to the maximum priority line and an alarm will be issued.
[0044] The present invention has at least the following beneficial effects:
[0045] The present invention firstly evaluates the stability of each type of flow data at each moment according to the fluctuation of each type of flow data at each moment and the same type of flow data in its neighborhood, and obtains the corresponding line stability value. Then, the comprehensive stability of each type of flow data at each moment in each sub-time period in the current time period is determined in combination with the difference between each type of flow data at each moment in each sub-time period in the current time period and the line stability value corresponding to the same type of flow data in the corresponding sub-time period in the historical time period. The sample point at each moment is obtained based on all types of flow data at each moment in the current time period. The distribution distance between different sample points is determined in combination with the difference in the comprehensive stability of each type of flow data at different sample points and the difference in the similarity of different types of flow data at different sample points and their neighborhoods. The present invention redefines the distribution distance between sample points in combination with the characteristics of the flow data of the line in the radio and television system, so that the distance between the sample points is more in line with the actual broadcasting status of the radio and television. Then, the CURE clustering algorithm is used for clustering, and the shrinkage factor is adaptively determined, so that the clustering is more flexible, the accuracy of clustering for flow conditions with different pressure levels is improved, and the flow distribution and alarm of the final line are more accurate and intelligent. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are 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.
[0047] Figure 1 A flowchart of a multi-line switching alarm method for intelligent traffic balancing control of broadcast television provided by an embodiment of the present invention;
[0048] Figure 2 A flowchart of a method for obtaining comprehensive stability;
[0049] Figure 3 Flow chart of the method for obtaining target clustering clusters. DETAILED DESCRIPTION
[0050] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the following is a detailed description of a multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television proposed by the present invention in combination with the accompanying drawings and preferred embodiments.
[0051] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0052] The specific scheme of the multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television provided by the present invention is described in detail below with reference to the accompanying drawings.
[0053] A multi-line switching alarm method embodiment for intelligent traffic balancing control of broadcast television:
[0054] The specific scenario targeted by this embodiment is: in view of the increasing traffic and signal processing requirements faced by the intelligent system of radio and television, this embodiment analyzes the operating conditions of multiple lines in real time, adjusts the shrinkage factor of the CURE clustering algorithm to better adapt to the distribution of sample data, and makes the final clustering more accurate and more adaptable, which facilitates the adjustment and switching of line traffic with abnormal operating conditions.
[0055] This embodiment proposes a multi-line switching alarm method for intelligent traffic balancing control of broadcast television, such as Figure 1 As shown, a multi-line switching alarm method for intelligent traffic balancing control of broadcast television in this embodiment includes the following steps:
[0056] Step S1, obtaining different types of traffic data of different lines in the broadcasting and television intelligent system at each moment in the current time period and each sub-time period in each historical time period.
[0057] In the intelligent system for broadcasting and television, the flow data of each line in the current time period is first collected. In this embodiment, two types of flow data are collected, namely bandwidth utilization data and transmission rate data. Among them, the bandwidth utilization data can reflect the load situation of the line, and then evaluate whether there is flow impact; the transmission rate data can be used to reflect the stability and speed of transmission, so as to determine whether there is flow impact; the current time period in this embodiment is a set of all historical moments whose time interval with the current moment is less than or equal to the preset time length. In this embodiment, the preset time length is 1 day. In specific applications, the implementer can set the preset time length and the type of flow data according to specific circumstances; at the same time, the flow data in each historical time period of each line in the intelligent system for broadcasting and television is obtained. The historical time period is the time period before the current time period. In this embodiment, the number of historical time periods is 10, and the length of each historical time period is 1 day. The flow data is collected once every second; in specific applications, the implementer can set the number of historical time periods and the frequency of flow data collection according to specific circumstances. The multi-line traffic data of radio and television is different in different time periods, because different types of program content have different dynamics and complexity, which will affect the size and transmission requirements of video stream data, so that different lines may correspond to different traffic performance conditions, and the content broadcast by radio and television in the same time period every day is highly similar, for example: 7:00-7:30, 12:00-12:30, 19:00-19:30, etc. are generally the time for news broadcasts. In view of this situation, the current time period and each historical time period are divided into multiple sub-time periods according to a fixed duration. In this embodiment, the duration of each sub-time period is 15 minutes. In specific applications, the implementer can set it according to the specific situation.
[0058] So far, this embodiment has obtained different types of traffic data for each line in the broadcasting and television intelligent system at each moment in the current time period and each sub-time period in each historical time period.
[0059] Step S2, according to the fluctuation of each type of traffic data at each moment and the same type of traffic data in its neighborhood, obtain the line stability value corresponding to each type of traffic data at each moment; according to the difference between the line stability values corresponding to each type of traffic data at each moment in each sub-time period in the current time period and the same type of traffic data in the corresponding sub-time period in the historical time period, and each type of traffic data at each moment in each sub-time period in each historical time period, obtain the comprehensive stability of each type of traffic data at each moment in each sub-time period in the current time period.
[0060] Since the program content has different dynamics and complexities, it will cause the traffic data to fluctuate. This embodiment will analyze the fluctuation trend of different types of traffic data and the impact of stable state performance. The closer the performance of different types of traffic data at different times is, the higher the similarity between these traffic data is. At the same time, the daily program schedule of radio and television is relatively consistent, that is, the rhythm of broadcasting and the content of broadcasting are relatively consistent. Therefore, this embodiment will determine the line stability value corresponding to each type of traffic data at each time according to the fluctuation of each type of traffic data at each time and the same type of traffic data in its neighborhood.
[0061] Step S21, determining the line stability value corresponding to each type of flow data at each moment according to the fluctuation of each type of flow data at each moment and the same type of flow data in its neighborhood.
[0062] Specifically, for the j-th type of traffic data at the i-th moment:
[0063] The j-th flow data at the ith moment and the j-th flow data at each moment in the neighborhood of the ith moment are subjected to curve fitting to obtain the curve corresponding to the j-th flow data at the ith moment; the neighborhood duration of this embodiment is 1 minute, so the method for obtaining the moments in the neighborhood of the ith moment is: all moments within the half minute before the ith moment and the half minute after the ith moment are taken as moments in the neighborhood of the ith moment. The method provided in this embodiment can be used to obtain the moments in the neighborhood of other moments. In specific applications, the implementer can set the neighborhood duration according to specific circumstances. According to the difference in data values between the extreme points on the curve corresponding to the j-th flow data at the ith moment, the line stability value corresponding to the j-th flow data at the ith moment is obtained. Curve fitting and the method for obtaining the extreme points on the curve are both prior arts and will not be elaborated here. The specific calculation formula for the line stability value corresponding to the j-th flow data at the ith moment is:
[0064]
[0065] Among them, Q i,jrepresents the line stability value corresponding to the j-th flow data at the ith moment, M represents the number of extreme points on the curve corresponding to the j-th flow data at the ith moment, Δq(m,m+2) represents the difference in data values between the m-th extreme point and the m+2-th extreme point on the curve corresponding to the j-th flow data at the ith moment, Δq(m+1,m+3) represents the difference in data values between the m+1-th extreme point and the m+3-th extreme point on the curve corresponding to the j-th flow data at the ith moment, σ(i,j) represents the standard deviation of the difference in data values between every two adjacent extreme points on the curve corresponding to the j-th flow data at the ith moment, norm() represents the normalization function, || represents the absolute value sign, λ1 is the preset first adjustment parameter, and λ1 is greater than 0.
[0066] The process of obtaining the difference of the data values between the two extreme value points is: taking the absolute value of the difference between the data values of the two extreme value points as the difference of the data values between the two extreme value points. It is used to reflect the difference between the difference between adjacent maximum values and the difference between adjacent minimum values, to reflect the data stability of horizontal comparison. The preset first adjustment parameter is introduced into the calculation formula of the line stability value to prevent the denominator from being 0. In this embodiment, the first adjustment parameter is 0.01. In specific applications, the implementer can set it according to the specific situation; σ(i,j) reflects the data stability during vertical comparison in the form of standard deviation; by combining horizontal and vertical, it comprehensively reflects the line stability and determines the line stability value corresponding to the jth flow data at the i-th moment. When the difference between the difference between adjacent maximum values and the difference between adjacent minimum values on the curve corresponding to the jth flow data at the i-th moment is smaller, and the standard deviation of the difference between the data values of each two adjacent extreme points on the curve corresponding to the jth flow data at the i-th moment is smaller, it means that the jth flow data at the i-th moment is more stable, that is, the line stability value corresponding to the jth flow data at the i-th moment is larger.
[0067] By adopting the above method, the line stability value corresponding to each type of flow data at each moment in the current time period and each historical time period can be obtained.
[0068] Step S22, combining the difference between each type of flow data at each moment in each sub-time period in the current time period and the line stability value corresponding to the same type of flow data in the corresponding sub-time period in the historical time period, obtains the comprehensive stability of each type of flow data at each moment in each sub-time period in the current time period.
[0069] The line stability value is determined based on the stability of the traffic data at each moment and in the neighborhood. For TV programs, the content and rhythm played in the same time interval on different dates are similar, so there should be a certain similarity between the stability of their traffic data. Therefore, based on the difference between the line stability values corresponding to each type of traffic data at each moment in each sub-time period in the current time period and the same type of traffic data in the corresponding sub-time period in the historical time period, and each type of traffic data at each moment in each sub-time period in each historical time period, the comprehensive stability of each type of traffic data at each moment in each sub-time period in the current time period is evaluated, such as Figure 2 As shown, the figure is a flow chart of the method for obtaining comprehensive stability.
[0070] Specifically, for the b-th moment in the k-th sub-time period in the current time period:
[0071] For the jth type of traffic data: the average value of the line stability value corresponding to the jth type of traffic data at all moments in the kth sub-time period in each historical time period is recorded as the first average value corresponding to the kth sub-time period in each historical time period; it should be noted that for the jth type of traffic data, each kth sub-time period in each historical time period corresponds to a first average value. The absolute value of the difference between the bth moment in the kth sub-time period in each historical time period and the corresponding first average value is recorded as the first difference corresponding to the kth sub-time period in each historical time period, and each kth sub-time period in each historical time period corresponds to a first difference. The absolute value of the difference between the line stability value corresponding to the jth type of traffic data at the bth moment in the kth sub-time period in the current time period and the line stability value corresponding to the jth type of traffic data at the bth moment in the kth sub-time period in each historical time period is recorded as the second difference corresponding to the bth moment in the kth sub-time period in each historical time period, and each kth sub-time period in the historical time period corresponds to a second difference. According to the first difference, the second difference and the line stability value corresponding to the bth moment in the kth sub-time period in the current time period, the comprehensive stability of the jth type of flow data at the bth moment in the kth sub-time period in the current time period is obtained. The specific calculation formula for the comprehensive stability of the jth type of flow data at the bth moment in the kth sub-time period in the current time period is:
[0072]
[0073] Among them, W k,b,j represents the comprehensive stability of the jth type of traffic data at the bth moment in the kth sub-time period in the current time period, F represents the number of historical time periods, Q′ k,b,j Indicates the line stability value corresponding to the j-th flow data at the b-th time in the k-th sub-time period in the current time period, It represents the average value of the line stability value corresponding to the j-th flow data at all times in the k-th sub-time period in the f-th historical time period, Q″ f,k,b,j It represents the line stability value corresponding to the j-th flow data at the b-th moment in the k-th sub-time period in the f-th historical time period. exp() represents an exponential function with a natural constant as the base, and || represents the absolute value sign.
[0074] is the first average value corresponding to the kth sub-time in the fth historical time period, It represents the first difference corresponding to the kth sub-time period in the fth historical time period, and analyzes the consistency between the stability of each historical flow data and the historical average stability. By taking the historical average stability as the standard, it reflects the degree to which the overall stability of the historical data approaches the historical average stability. The smaller the value, the more consistent the overall stability of the historical flow data. It indicates the second difference corresponding to the bth moment in the kth sub-time period in the fth historical time period, reflecting the consistency between the stability corresponding to the jth type of flow data at the bth moment in the kth sub-time period in the current time period and the historical average stability. The smaller the value, the closer the stability corresponding to the jth type of flow data at the bth moment in the kth sub-time period in the current time period is to the historical average stability. As the adjustment coefficient of the line stability value, it is used to adjust the line stability value corresponding to the j-th flow data at the b-th moment in the k-th sub-time period in the current time period to determine the comprehensive stability of the j-th flow data at the b-th moment in the k-th sub-time period in the current time period.
[0075] By adopting the above method, the comprehensive stability of each type of flow data at each moment in each sub-time period in the current time period can be obtained.
[0076] Step S3, based on all types of traffic data at each moment in the current time period, obtain the sample points at each moment; according to the difference in the comprehensive stability of each type of traffic data at different sample points and the difference in the similarity of different types of traffic data at different sample points and their neighborhoods, obtain the distribution distances between different sample points; based on the distribution distances, use the CURE clustering algorithm to cluster all sample points. In the clustering process, the clusters with a number of sample points greater than a preset number are recorded as initial clusters, and the shrinkage factor of each initial cluster is obtained according to the relative distance of the sample points in each initial cluster. Clustering is continued based on the shrinkage factor to obtain each target cluster.
[0077] This embodiment will determine the corresponding sample points by combining all types of traffic data at each moment in the current time period, and then use the CURE clustering algorithm to cluster all sample points, and adaptively determine the shrinkage factor to obtain accurate clustering results.
[0078] Step S31, obtaining a sample point at each moment based on all types of traffic data at each moment in the current time period.
[0079] For each moment in the current time period: in the coordinate system, use the time as the horizontal coordinate, the bandwidth utilization data as the vertical coordinate, and the transmission rate data as the vertical coordinate to obtain the data point of each moment and record it as the sample point of each moment.
[0080] Step S32, according to the difference in comprehensive stability of each type of flow data at different sample points and the difference in similarity of different types of flow data at different sample points and their neighborhoods, the distribution distance between different sample points is obtained, and the CURE clustering algorithm is used to cluster all sample points based on the distribution distance to obtain an initial cluster cluster.
[0081] Specifically, for the p-th sample point and the q-th sample point:
[0082] For the j-th type of traffic data: the difference in comprehensive stability between the j-th traffic data of the p-th sample point and the q-th sample point is recorded as the third difference corresponding to the j-th traffic data of the p-th sample point and the q-th sample point; the j-th traffic data of the p-th sample point and the j-th traffic data at all times in the neighborhood of the time corresponding to the p-th sample point constitute the j-th traffic data sequence corresponding to the p-th sample point, and the j-th traffic data of the q-th sample point and the j-th traffic data at all times in the neighborhood of the time corresponding to the q-th sample point constitute the j-th traffic data sequence corresponding to the q-th sample point; it should be noted that the traffic data in the traffic data sequence are arranged in the order of collection time. Calculate the arithmetic square root of the sum of the squares of the third differences corresponding to all types of traffic data between the p-th sample point and the q-th sample point; obtain the distribution distance between the p-th sample point and the q-th sample point based on the difference between the arithmetic square root and the correlation between each two traffic data sequences corresponding to the p-th sample point and the correlation between each two traffic data sequences corresponding to the q-th sample point, and the difference between the arithmetic square root and the correlation between each two traffic data sequences is positively correlated with the distribution distance. Among them, the positive correlation relationship means that the dependent variable will increase as the independent variable increases, and the dependent variable will decrease as the independent variable decreases. The specific relationship can be a multiplication relationship, an addition relationship, and the idempotence of an exponential function, which is determined by actual application; as a specific implementation method, a specific calculation formula for the distribution distance is given. Since there are two types of traffic data collected in this embodiment, the specific calculation formula for the distribution distance between the p-th sample point and the q-th sample point is:
[0083]
[0084] Among them, D p,q represents the distribution distance between the pth sample point and the qth sample point, represents the difference in comprehensive stability between the first type of flow data at the p-th sample point and the q-th sample point, It represents the difference in comprehensive stability between the second flow data of the p-th sample point and the q-th sample point, It represents the difference between the correlation between each two traffic data sequences corresponding to the p-th sample point and the correlation between the first traffic data sequence and the second traffic data sequence corresponding to the q-th sample point.
[0085] is the third difference between the first type of flow data at the p-th sample point and the q-th sample point, is the third difference corresponding to the second type of traffic data of the pth sample point and the qth sample point. There is a certain correlation between bandwidth utilization and transmission rate. When the correlation difference corresponding to the two sample points is smaller, it means that the two maintain a relatively consistent balance in the corresponding relationship between bandwidth utilization and transmission rate. Therefore, the distance between the two should be reduced. This embodiment reflects the distance between the two sample points through the Euclidean distance of the comprehensive stability of the bandwidth utilization data and the comprehensive stability of the transmission rate data between the two sample points, and adjusts the distance through the correlation between the traffic data to determine the distribution distance between the two sample points.
[0086] It should be noted that: the present embodiment uses correlation to reflect the similarity between two sequences. The greater the correlation, the more similar the two sequences are. As other implementations, the cosine similarity between the two sequences may also be calculated, and the cosine similarity may be used to reflect the similarity between the two sequences. The greater the cosine similarity, the more similar the two sequences are. Alternatively, the negative correlation value of the dynamic time warping distance between the two sequences may be used to reflect the similarity between the two sequences. The smaller the dynamic time warping distance between the two sequences, the greater the negative correlation value of the dynamic time warping distance, and the more similar the two sequences are. As other implementations, if the types of traffic data collected are greater than two, the mean of the differences between the correlations may be used to replace the distribution distance calculation formula in the present embodiment. Then the calculation is performed to obtain the distribution distance.
[0087] By adopting the above method, the distribution distance between every two sample points can be obtained. Next, this embodiment uses the CURE clustering algorithm to cluster all sample points based on the distribution distance between every two sample points. In the clustering process, the clusters with more sample points than the preset number are recorded as the initial clustering clusters. That is, in the clustering process, when the number of sample points in the cluster is greater than the preset number, it is used as the initial clustering cluster, and the shrinkage factor of the initial clustering cluster is adaptively adjusted later. The preset number in this embodiment is 10. In specific applications, the implementer can set it according to the specific situation.
[0088] Step S33, obtaining a shrinkage factor of each initial cluster according to the relative distance of the sample points in each initial cluster, and continuing clustering based on the shrinkage factor to obtain each target cluster.
[0089] When the CURE clustering algorithm is used to cluster sample points, when the number of sample points in the cluster reaches a preset number, this embodiment will process the sample representative points selected by the algorithm itself through the shrinkage factor, and adapt the shrinkage factor to analyze the density distribution in the initial cluster. When the density distribution of the initial cluster is larger, it means that the representative points are closer to the cluster center point, which can better reflect the overall performance of the current cluster. Based on this, this embodiment will adaptively determine the shrinkage factor of each initial cluster according to the relative distance of the sample points in each initial cluster.
[0090] For the cth initial cluster: calculate the cumulative sum of the distances between each sample point in the cth initial cluster and all other sample points in the cth initial cluster as the distance evaluation value of each sample point in the cth initial cluster; each sample point in the cth initial cluster corresponds to a distance evaluation value, and the sample point corresponding to the minimum distance evaluation value is used as the target point. Calculate the average distance between all sample points in the cth initial cluster and the target point; use the circular area with the target point as the center and the average distance as the radius as the reference area corresponding to the cth initial cluster; according to the difference between the distance between each sample point in the cth initial cluster and the target point and the average distance, the average distance, and the number of sample points in the reference area, obtain the shrinkage factor of the cth initial cluster. The specific calculation formula for the shrinkage factor of the cth initial cluster is:
[0091]
[0092] Among them, R c represents the shrinkage factor of the cth initial cluster, N represents the number of sample points in the cth initial cluster, represents the average distance between all sample points and the target point in the cth initial cluster, d c,n represents the distance between the nth sample point and the target point in the cth initial cluster, N′ represents the number of sample points in the reference area corresponding to the cth initial cluster, norm() represents the normalization function, and || takes the absolute value sign.
[0093] Indicates the difference between the distance between the nth sample point and the target point in the cth initial cluster and the average distance. It is used to reflect the overall difference between the distance and the average distance between the sample points and the target points in the cth initial cluster. When the average distance between all sample points and the target points in the cth initial cluster is larger, the overall difference between the distance and the average distance between the sample points and the target points in the cth initial cluster is larger, and the ratio of the number of sample points in the reference area corresponding to the cth initial cluster to the number of sample points in the cth initial cluster is larger, it means that the distribution of sample points in the cth initial cluster is less dense, and the shrinkage factor should be smaller, that is, the degree of shrinkage is greater.
[0094] By adopting the above method, the shrinkage factor of each initial cluster can be obtained. Based on the shrinkage factor of each initial cluster, the CURE clustering algorithm is continued to cluster the sample points until the number of clusters finally obtained reaches the preset requirement. Each cluster obtained after the final clustering is completed is recorded as a target cluster, that is, multiple target clusters are obtained. The preset requirement in this embodiment is 15. In specific applications, the implementer can set it according to the specific situation. Clustering based on the shrinkage factor using the CURE clustering algorithm is a prior art and will not be described in detail here.
[0095] Thus, the method provided in this embodiment is used to adaptively determine the shrinkage factor corresponding to the sample point in the clustering process, and then cluster all the sample points to obtain multiple target clusters, such as Figure 3 As shown, the figure is a flow chart of a method for obtaining target clustering clusters.
[0096] Step S4, judging whether to perform line switching and alarm based on the number of sample points in each target cluster.
[0097] In step S3, the present embodiment clusters the sample points corresponding to each moment in the current time period to obtain a plurality of target clusters. Next, it is determined whether to switch the line based on the number of sample points in the target clusters.
[0098] Specifically, the number of sample points in each target cluster is counted, and the number of sample points in the target cluster with the least number of sample points is recorded as the first number, and the number of sample points in the target cluster with the largest number of sample points is recorded as the second number; if the ratio of the first number to the second number is greater than or equal to the preset proportion threshold, the line is not switched and no alarm is issued; if the ratio of the first number to the second number is less than the preset proportion threshold, an alarm is issued and the corresponding line is adjusted to the maximum priority line. The priority of each line in the broadcasting and television intelligent system is set by the implementer according to the specific situation. In this embodiment, the preset proportion threshold is 0.05. In specific applications, the implementer can set it according to the specific situation.
[0099] At this point, the method provided in this embodiment has been used to complete the balanced control of the broadcasting and television intelligent traffic.
[0100] This embodiment firstly evaluates the stability of each type of flow data at each moment according to the fluctuation of each type of flow data at each moment and the same type of flow data in its neighborhood, and obtains the corresponding line stability value. Then, combined with the difference between each type of flow data at each moment in each sub-time period in the current time period and the line stability value corresponding to the same type of flow data in the corresponding sub-time period in the historical time period, the comprehensive stability of each type of flow data at each moment in each sub-time period in the current time period is determined. Based on all types of flow data at each moment in the current time period, the sample points at each moment are obtained. Combined with the difference in the comprehensive stability of each type of flow data at different sample points and the difference in the similarity of different types of flow data at different sample points and their neighborhoods, the distribution distance between different sample points is determined. This embodiment redefines the distribution distance between sample points in combination with the characteristics of the flow data of the line in the radio and television system, so that the distance between sample points is more in line with the actual broadcasting of radio and television. Then, the CURE clustering algorithm is used for clustering, and the shrinkage factor is adaptively determined to make clustering more flexible, improve the accuracy of clustering for flow conditions with different pressure levels, and make the final line flow distribution and alarm more accurate and intelligent.
[0101] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television, characterized in that: The method comprises the following steps: Obtain different types of traffic data for different lines in the radio and television intelligent system at each moment in the current time period and each sub-time period in each historical time period; According to the fluctuation of each type of traffic data at each moment and the same type of traffic data in its neighborhood, the line stability value corresponding to each type of traffic data at each moment is obtained; according to the difference between each type of traffic data at each moment in each sub-time period in the current time period and the line stability value corresponding to the same type of traffic data in the corresponding sub-time period in the historical time period, and each type of traffic data at each moment in each sub-time period in each historical time period, the comprehensive stability of each type of traffic data at each moment in each sub-time period in the current time period is obtained; Based on all kinds of traffic data at each moment in the current time period, sample points at each moment are obtained; according to the difference in comprehensive stability of each kind of traffic data at different sample points and the difference in similarity of different kinds of traffic data at different sample points and their neighborhoods, distribution distances between different sample points are obtained; based on the distribution distances, all sample points are clustered using the CURE clustering algorithm; in the clustering process, clusters with a number of sample points greater than a preset number are recorded as initial clusters; a contraction factor of each initial cluster is obtained based on the relative distance of sample points within each initial cluster; clustering is continued based on the contraction factor to obtain each target cluster; Determine whether to switch lines and issue an alarm based on the number of sample points in each target cluster; The determining whether to perform line switching and alarm based on the number of sample points in each target cluster includes: The number of sample points in the target cluster with the least number of sample points is recorded as the first number, and the number of sample points in the target cluster with the largest number of sample points is recorded as the second number; If the ratio of the first number to the second number is greater than or equal to the preset proportion threshold, the line will not be switched and no alarm will be issued; if the ratio of the first number to the second number is less than the preset proportion threshold, the corresponding line will be adjusted to the maximum priority line and an alarm will be issued.
2. According to claim 1, a multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television is characterized in that: The method of obtaining the line stability value corresponding to each type of flow data at each moment according to the fluctuation of each type of flow data at each moment and the same type of flow data in its neighborhood includes: For the j-th type of traffic data at the i-th moment: Perform curve fitting on the j-th flow data at the i-th moment and the j-th flow data at each moment in the neighborhood of the i-th moment to obtain the curve corresponding to the j-th flow data at the i-th moment; According to the difference in data values between extreme points on the curve corresponding to the j-th type of flow data at the i-th moment, the line stability value corresponding to the j-th type of flow data at the i-th moment is obtained.
3. A multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television according to claim 2, characterized in that: The line stability value corresponding to the j-th flow data at the i-th moment is calculated using the following formula: Among them, Q i,j represents the line stability value corresponding to the j-th flow data at the ith moment, M represents the number of extreme points on the curve corresponding to the j-th flow data at the ith moment, Δq(m,m+2) represents the difference in data values between the m-th extreme point and the m+2-th extreme point on the curve corresponding to the j-th flow data at the ith moment, Δq(m+1,m+3) represents the difference in data values between the m+1-th extreme point and the m+3-th extreme point on the curve corresponding to the j-th flow data at the ith moment, σ(i,j) represents the standard deviation of the difference in data values between every two adjacent extreme points on the curve corresponding to the j-th flow data at the ith moment, norm() represents the normalization function, || represents the absolute value sign, λ1 is the preset first adjustment parameter, and λ1 is greater than 0.
4. A multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television according to claim 1, characterized in that: The comprehensive stability of each type of flow data at each moment in each sub-time period in the current time period is obtained according to the difference between each type of flow data at each moment in each sub-time period in the current time period and the line stability value corresponding to the same type of flow data in the corresponding sub-time period in the historical time period, and each type of flow data at each moment in each sub-time period in each historical time period, including: For the j-th type of traffic data at the b-th moment in the k-th sub-time period of the current time period: The average value of the line stability value corresponding to the j-th flow data at all moments in the k-th sub-time period in each historical time period is recorded as the first average value corresponding to the k-th sub-time period in each historical time period; The absolute value of the difference between the b-th moment in the k-th sub-time period in each historical time period and the corresponding first average value is recorded as the first difference corresponding to the k-th sub-time period in each historical time period; The absolute value of the difference between the line stability value corresponding to the j-th flow data at the b-th moment in the k-th sub-time period in the current time period and the line stability value corresponding to the j-th flow data at the b-th moment in the k-th sub-time period in each historical time period is recorded as the second difference corresponding to the b-th moment in the k-th sub-time period in each historical time period; According to the first difference, the second difference and the line stability value corresponding to the bth moment in the kth sub-time period in the current time period, the comprehensive stability of the jth type of flow data at the bth moment in the kth sub-time period in the current time period is obtained.
5. A multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television according to claim 4, characterized in that: The following formula is used to calculate the comprehensive stability of the j-th flow data at the b-th time in the k-th sub-time period in the current time period: Among them, W k,b,j represents the comprehensive stability of the jth type of traffic data at the bth moment in the kth sub-time period in the current time period, F represents the number of historical time periods, Q′ k,b,j Indicates the line stability value corresponding to the j-th flow data at the b-th time in the k-th sub-time period in the current time period, It represents the average value of the line stability value corresponding to the j-th flow data at all times in the k-th sub-time period in the f-th historical time period, Q″ f,k,b,j It represents the line stability value corresponding to the j-th flow data at the b-th moment in the k-th sub-time period in the f-th historical time period. exp() represents an exponential function with a natural constant as the base, and || represents the absolute value sign.
6. A multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television according to claim 1, characterized in that: The distribution distances between different sample points are obtained based on the differences in the comprehensive stability of each type of flow data at different sample points and the differences in the similarity of different types of flow data at different sample points and their neighborhoods, including: For the p-th sample point and the q-th sample point: For the jth type of traffic data: the difference in comprehensive stability between the jth type of traffic data of the pth sample point and the qth sample point is recorded as the third difference corresponding to the jth type of traffic data of the pth sample point and the qth sample point; the jth type of traffic data of the pth sample point and the jth type of traffic data at all times in the neighborhood of the time corresponding to the pth sample point constitute the jth type of traffic data sequence corresponding to the pth sample point, and the jth type of traffic data of the qth sample point and the jth type of traffic data at all times in the neighborhood of the time corresponding to the qth sample point constitute the jth type of traffic data sequence corresponding to the qth sample point; According to the third difference corresponding to each flow data of the p-th sample point and the q-th sample point, and the difference between the correlation between each two flow data sequences corresponding to the p-th sample point and the correlation between each two flow data sequences corresponding to the q-th sample point, the distribution distance between the p-th sample point and the q-th sample point is obtained.
7. A multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television according to claim 6, characterized in that: The method of obtaining the distribution distance between the p-th sample point and the q-th sample point according to the third difference corresponding to each flow data of the p-th sample point and the q-th sample point, and the difference between the correlation between each two flow data sequences corresponding to the p-th sample point and the correlation between each two flow data sequences corresponding to the q-th sample point, includes: Calculate the arithmetic square root of the sum of squares of the third differences corresponding to all types of traffic data of the p-th sample point and the q-th sample point; According to the difference between the correlation between the arithmetic square root and each two traffic data sequences corresponding to the p-th sample point and the correlation between each two traffic data sequences corresponding to the q-th sample point, the distribution distance between the p-th sample point and the q-th sample point is obtained, and the difference between the arithmetic square root and the correlation between each two traffic data sequences is positively correlated with the distribution distance.
8. A multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television according to claim 1, characterized in that: The step of obtaining a shrinkage factor of each initial cluster according to the relative distance of the sample points in each initial cluster, and continuing clustering based on the shrinkage factor to obtain each target cluster includes: For the cth initial cluster: Calculate the cumulative sum of the distances between each sample point in the cth initial cluster and all other sample points in the cth initial cluster as the distance evaluation value of each sample point in the cth initial cluster; take the sample point corresponding to the minimum distance evaluation value as the target point; Calculate the average distance between all sample points and the target point in the cth initial cluster; take a circular area with the target point as the center and the average distance as the radius as the reference area corresponding to the cth initial cluster; A shrinkage factor of the c th initial cluster is obtained according to the difference between the distance between each sample point in the c th initial cluster and the target point and the average distance, the average distance, and the number of sample points in the reference area.
9. A multi-line switching alarm method for intelligent traffic balancing control of broadcasting and television according to claim 8, characterized in that: The shrinkage factor of the cth initial cluster is calculated using the following formula: Among them, R c represents the shrinkage factor of the cth initial cluster, N represents the number of sample points in the cth initial cluster, represents the average distance between all sample points and the target point in the cth initial cluster, d c,n represents the distance between the nth sample point and the target point in the cth initial cluster, N′ represents the number of sample points in the reference area corresponding to the cth initial cluster, norm() represents the normalization function, and || takes the absolute value sign.
