Big data and artificial intelligence telegraph management analysis system

Abstract

The invention provides a telegraph management analysis system based on big data and artificial intelligence, and belongs to the field of telegraph management. The problem of low telegraph management efficiency is solved; specifically, the telegraph acquisition module is used for acquiring a telegraph to be processed; the service management module is used for demodulating and backing up the telegraph to be processed to obtain a quasi telegraph; employee information and historical telegraph data are obtained, duty arrangement and duty arrangement adjustment are carried out on operators on duty, and a quasi telegraph is sent; the telegraph storage sub-module is used for analyzing and extracting a telegraph head and a telegraph tail of a quasi telegraph, establishing a storage structure and storing the quasi telegraph; analyzing the electric signal change of the quasi telegraph, constructing a telegraph equation, and sending an electric signal enhancement instruction to a subordinate telegraph station; through acquisition, processing and mathematical modeling of related data of the telegraph, the receiving quantity and the sending quantity of the telegraph every day are analyzed, the on-duty sequence is optimized, the telegraph is stored, and the electric signal strength of the telegraph is adjusted, so that the telegraph can smoothly reach a target unit.

Description

Technical Field

[0001] The present invention discloses a big data and artificial intelligence telegram management and analysis system, which relates to the field of telegram management. Background Art

[0002] The existing telegram management system has the following deficiencies: High technical and resource requirements: The existing telegram management and analysis system requires high-performance computing equipment, advanced software tools and a large amount of storage space to support its operation; secondly, in terms of software, the system needs to adopt advanced data analysis tools and algorithm libraries, etc. These software tools are expensive and need to be constantly updated and upgraded to maintain their competitiveness, which is not conducive to the long-term operation of the enterprise.

[0003] Shortage of technical personnel: Since big data analysis and artificial intelligence technologies involve complex statistical and mathematical models, their analysis results will be affected by many factors. It is difficult for general telegraph business personnel to understand and apply these analysis results. Some statistical or physics technicians are needed to identify and screen them, which limits the application effect of big data analysis in telegraph business decision-making.

[0004] Complex management: In terms of telegram management, enterprises need to establish a complete telegram management system, including the collection, storage, processing and analysis of telegrams; at the same time, they also need to formulate strict telegram management processes, which makes the integration and management of telegrams complicated. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a telegram management and analysis system based on big data and artificial intelligence, aiming to solve the problem of low efficiency of telegram management.

[0006] In order to achieve the above object, the present invention is implemented through the following technical solutions: A big data and artificial intelligence telegram management and analysis system includes: Telegram acquisition module: used to obtain received telegrams and obtain telegrams to be processed; Business management module: used to demodulate and back up the telegram to be processed to obtain the quasi-telegram; encrypt and modulate the quasi-telegram to obtain the telegram to be sent; Get the number of on-duty personnel, the number of telegrams sent by each on-duty personnel, and obtain employee information; Obtain the amount of telegrams received and sent in the past three months to obtain historical telegram data; perform time series analysis on historical telegram data to calculate the expected receiving rate and expected sending rate of telegram sending work; obtain duty requirements, and schedule and adjust the duty personnel in combination with employee information, and send the telegrams to be sent; The telegram processing module includes: a telegram storage submodule and a telegram analysis submodule: Telegram storage submodule: used to parse and extract the telegram header and telegram tail of the quasi-telegram, obtain the receiving information and confidentiality level of the quasi-telegram; perform secondary encryption on the quasi-telegram according to the confidentiality level, and establish a storage structure according to the receiving information of the quasi-telegram to save the quasi-telegram; Telegraph analysis submodule: used to analyze the changes in the electrical signals of the quasi-telegraph and construct telegraph equations; according to the telegraph equations, electrical signal enhancement instructions are sent to the lower-level telegraph stations until the quasi-telegraph is received by the target unit.

[0007] Furthermore, the workflow of the business management module is as follows: Process A1: Duty requirements include: daily working hours, duty days, number of staff on duty in a single shift, and maximum number of shifts for a single person on duty; Count the number of telegrams to be processed te; demodulate and back up the telegrams to be processed to obtain quasi-telegrams; encrypt and modulate the quasi-telegrams to obtain telegrams to be sent; Process A2: Get the number of on-duty personnel st; get the number of telegrams received in the past three months re （1） ~re （td） ; Telegram sending volume （1） ~se （td） ; Among them, td represents the total number of days in the past three months; Get the number of days dd that the 1st to the st duty personnel have been on duty （1） ~dd （st） , the amount of telegrams sent sn （1） ~sn （st） ; Process A3: Get the daily working time wh; define calculation formula A-1: ; Among them, sn （i） ,dd （i） andve （i） , respectively represent the number of telegrams sent by the i-th duty officer, the number of days on duty, and the rate of telegrams sent; Calculate the telegram sending rate of the 1st to the st-th duty personnel, and get ve （1） ~ve （st） ; Process A4: Re （1） ~re （td） and （1） ~se （td） , respectively conduct time series analysis and construct the telegram receiving amount prediction equation and the telegram sending amount prediction equation; Process A5: Get the differential order d, and get the telegram reception volume rv in the previous d days （1） ~rv （d） , telegram sending volume sd （1） ~sd （d） ; rv （1） ~rv （d） and sd （1） ~sd （d） , substitute them into the telegram reception prediction equation and the telegram sending prediction equation respectively, and calculate the total telegram reception volume zrv and the total telegram sending volume zse for a single day.

[0008] Furthermore, the subsequent process of process A5 is as follows: Process A6: record the number of on-duty personnel in a single shift as pp, the number of days on duty as wd, and the maximum number of shifts on duty by a single person as ww; Definition of the mathematical model of the duty officer A-6-1: ; Among them, pe （i,j） It means that the i-th duty officer is on duty on the j-th day; Get the current working day we; Set the value of j to we and set pe （i,j） The corresponding number of telegrams sent by the class personnel is denoted as ve （i,j） ; Define constraint A-6-2: ; Define the objective function A-6-3: ; Where ZZ represents the cost; Process A7: According to A-6-2 and A-6-3, use the linear programming solver to solve the mathematical model A-6-1 and obtain the linear combination with the minimum cost as the initial duty sequence; Count the number of people pr in the initial duty sequence; obtain the number of telegrams sent by the 1st to prth scheduled personnel, recorded as vee （1） ~vee （pr） ; Calculate vee （1） ~vee （pr） of and ave; Get yesterday's telegram reception volume hrv; get the previous (d+1) day's telegram reception volume rv （d+1） ; rv （2） ~rv （d+1） Substitute into the telegram sending volume prediction equation to calculate yesterday's expected telegram receiving volume qrv; where rv （2） It indicates the amount of telegrams received in the previous 2 days; Process A8: Determine whether hrv≤qrv is true; If established, the initial duty sequence remains unchanged; If it does not hold, change constraint A-6-2 to constraint A-6-4: ; Replace A-6-2 with A-6-4 and re-execute process A7 to obtain the optimal duty sequence; Process A9: Arrange the duty personnel according to the duty order of the optimal duty sequence and send the telegrams to be sent.

[0009] Furthermore, the specific process of process A4 is as follows: Process A41: The telegram reception amount prediction equation is used as equation A; according to re （1） ~re （td） Construct equation A; Process A42: Re （1） ~re （td） Replace with se （1） ~se （td） , repeat the same process of constructing equation A to construct the telegram sending volume prediction equation.

[0010] Furthermore, the specific process of the process A41 is as follows: Process A411: （1） ~re （td）} as the time series Yt; use ADF detection to perform difference detection on Yt and obtain the difference order d; Taking d-order differences on Yt, we get ▽ （d） Yt; processed using autocorrelation function and partial autocorrelation function▽ （d） Yt, get the autoregressive parameter p and sliding average parameter q of Yt; Process A412: Calculate re （1） ~re （td） The average values ​​are; Define calculation formula A-2: ;re （x） and ε （x） , respectively represent the amount of telegrams received and white noise on the xth day; Calculate the white noise from day 1 to day td: ε （1） ~ε （td） ; Set {ε （1） ~ε （td）} as a white noise sequence εt; Process A413: Construct an initial expression of equation A as formula A-3: ; Where, k and l both represent time difference; B （k） represents the lag operator of Yt when the time difference is k; BB （l） represents the lag operator of εt when the time difference is l; φ （k） Indicates B （k）The autocorrelation coefficient, θ （l） Indicates BB （l） is the sliding average coefficient of , and σ represents the error term.

[0011] Furthermore, the subsequent process of the process A413 is as follows: Process A414: Reset the amount of telegrams received on day x （x） Substituting into formula A-3 and expanding it, we get formula A-4: ; in, Represents （1） To （x） The d-order difference of Represents （1） To （x-1） The d-order difference of Represents （1） To （x-p） The d-order difference of φ （1） ~φ （p） represents the 1st to pth order autocorrelation coefficient; ε （x） represents the white noise of the xth day, ε （x-1） represents the white noise of the (x-1)th day; similarly, ε （x-q） represents the white noise of the (x-q)th day; θ （1） ~θ （q） represents the 1st to qth order sliding mean coefficient; φ （1） ~φ （p） and θ （1） ~θ （q） The initial value of is 1; Process A415: Extract the left part of formula A-4 and calculate φ （1） ~φ （p） The specific value of The left side of formula A-4 is: ; Process A416: Extract the right part of formula A-4 and repeat the calculation of φ （1） ~φ （p） The same process as the specific value is used to calculate θ （1） ~θ （q） The specific value of The right side of formula A-4 is: ; φ （1） ~φ （p） The specific value of θ （1） ~θ（q） Substituting the specific value of into formula A-4, we can obtain the telegram reception prediction equation, namely equation A.

[0012] Furthermore, the specific process of the process A415 is as follows: Process A4151: First-order difference; for re （1） ~re （td） Perform first-order difference to obtain the difference value ri （1） ~ri （tf） ; Where tf represents the number of samples after the first-order difference, tf = td-1; Calculate ri （1） ~ri （tf） The corresponding white noise cε （1） ~cε （tf） ; Process A4152: In rc （1+p） ~rc （tf） Take any p consecutive subscript values ​​in as sample values, denoted as rc （sa） ~rc （sa+p） ; {rc （sa） ~rc （sa+p）} as set A; calculate the sample value of set A lagged by 1 day, and get set A （1） Similarly, calculate the sample value of the lag p days and get the set A （p） ; Process A4153: Calculate rc （sa） ~rc （sa+p） The variance of is denoted as σ(rc); According to set A and set A （1） ~Set A （p） , calculate the covariance of the sample values ​​lagged 1 to p days respectively, and get γ(rc （1） )~γ(rc （p） ); Define calculation formula A-5-1: rr （k） =γ(rc （k） ) / σ(rc);γ(rc （k） ) and rr （k） , respectively represent the covariance and correlation coefficient of the sample value lagged k days; Calculate the correlation coefficient rr of the sample value lagged 1 to p days （1） ~rr （p） .

[0013] Furthermore, the subsequent process of process A4153 is as follows: Process A4154: φ （1） ~φ （p） The first-order difference estimate of（1，1） ~φ （1，p） ; Construct an empty matrix of (p×1) and fill it with φ （1，1） ~φ （1，p） and rr （1） ~rr （p） , we get matrix AA and matrix AB; The matrix AA is: ; Matrix AB is: ; Among them, φ （1，2） Represents φ （2） The first-order difference estimate of , rr （2） Correlation coefficient of sample values ​​lagged 2 days; Process A4155: Determine whether (2×p-2)≤tf holds and define the calculation formula of AA; If not, the calculation formula of AA is: AA = (AB × AB T ) -1 ×AB; where × represents matrix multiplication, T represents the transpose of the matrix, and -1 represents the inverse of the matrix; If it is established, repeat process A4152 to process A4153 to calculate the correlation coefficient rr of the sample value lagged (-1) to [-(p-2)] days （p+1） ~rr （2×p-2） ; Construct an empty matrix of (p×p) and fill it with σ(rc) and rr respectively. （1） ~rr （2×p-2） , we get the matrix AC; ; Among them, rr （2） ,rr （3） ,rr （4） and rr （p-1） , respectively represent the correlation coefficients of sample values ​​lagged by 2 days, 3 days, 4 days, and (p-1) days; The calculation formula of AA is: AA = (AC) -1 ×AB.

[0014] Furthermore, the subsequent process of process A4155 is as follows: Process A4156: Calculate φ （1，1） ~φ （1，p） The value of （1） ~qφ （p） ; Let the yth difference value be ri （y） ,ri （y） The corresponding white noise is cε （y） ; The value range of y is: (p+1)~tf; Construction formula A-5-2: ; where φ （1，k） represents the first-order difference estimate of the k-order autocorrelation coefficient, qφ （k） Represents φ （1，k） The specific value of ri （y-k） Represents （y） The difference value lagged k days; Represents （y） The estimated value of; the value range of k is: 1~p; Will ri （p+1） ~ri （tf） Substitute into formula A-5-2 to test and determine ri （p+1） ~ri （tf） The corresponding estimated value is （p+1） ~ri （tf） Whether the difference is within ξ; If yes, keep φ （1，1） ~φ （1，p） ; If not, modify φ （1，1） ~φ （1，p） The value of is taken, and the process A4151 to process A4156 are repeatedly executed until the test of formula A-5-2 is passed; wherein ξ represents the error determination coefficient; Process A4157: Repeated calculation of qφ （1） ~qφ （p） The process of calculating φ （1） ~φ （p） The 1st to pth order autocorrelation coefficient in the second order difference: φ （2，1） ~φ （2，p） ; Similarly, the 1st to pth order autocorrelation coefficient of the d-order difference is: φ （d，1） ~φ （d，p） ; Calculate qφ （1） and φ （2，1） ~φ （d，1） The average value of φ （1） The value of Similarly, calculate qφ （p） and φ （d，p） ~φ （d，p） The average value of φ （p） The value of .

[0015] Compared with the prior art, the present invention has the following beneficial effects: Intelligent business management: The present invention acquires, analyzes and mathematically models the historical telegram reception data and historical telegram transmission data of the telegraph station, and then formulates a shift sequence based on the business capabilities of the on-duty personnel, thereby saving manpower to the maximum extent while ensuring the telegram transmission task.

[0016] Trend analysis: The present invention can predict future communication needs and trends by analyzing historical data, and provide more scientific and reasonable resource allocation suggestions to relevant organizations, which helps to reduce operating costs and improve resource utilization efficiency.

[0017] Promoting information sharing: The present invention can establish storage structures for different telegrams according to the received information and confidentiality level of the telegrams, so as to facilitate information sharing and communication between different departments and teams; at the same time, the present invention can also analyze the physical properties of telegram signals, send signal enhancement instructions to lower-level telegraph stations, and ensure stable transmission of telegrams. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 It is a schematic diagram of the system of the present invention; Figure 2 It is a schematic diagram of the telegram of the present invention; Figure 3 This is a schematic diagram of the Merkle tree of the present invention. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] See also Figure 1 and Figure 2 , a big data and artificial intelligence telegram management and analysis system includes: Telegram acquisition module: used to obtain the received telegrams (on the same day) and get the telegrams to be processed; Business management module: used to demodulate and back up the telegram to be processed to obtain the quasi-telegram; encrypt and modulate the quasi-telegram to obtain the telegram to be sent; Get the number of on-duty personnel (at the telegraph station), the number of telegrams sent by each on-duty personnel (in the past three months), and obtain employee information; Obtain the number of telegrams received and sent by (telegraph station) in the past three months to obtain historical telegram data; perform time series analysis on historical telegram data to calculate the expected reception rate and expected sending rate of telegram sending work; obtain duty requirements, and schedule and adjust the duty personnel in combination with employee information, and send the telegrams to be sent; Duty requirements include: (within a week) daily working hours, number of duty days, number of personnel on duty in a single shift, and the maximum number of shifts for a single person on duty; Process A: The workflow of the business management module is as follows: Process A1: record the current time as (mo month da day) ho hour; count the number of telegrams to be processed, record it as te; The telegram to be processed is demodulated and backed up to obtain a quasi-telegram; the quasi-telegram is encrypted and modulated to obtain a telegram to be sent; (the number of telegrams to be sent is also te) Process A2: Get the number of on-duty personnel, recorded as st; Get the number of telegrams received by (telegraph station) in the past three months, recorded as re （1） ~re （td） ; The amount of telegrams sent, denoted as se （1） ~se （td） ; Among them, td represents the total number of days in the past three months; Get the number of days the 1st to the stth duty personnel have been on duty in the past three months, recorded as dd （1） ~dd （st） ; (where 1≤{dd （1） ~dd （st）}≤td, that is, each on-duty personnel has been on duty for at least 1 day and at most td days in the past three months) Get the number of telegrams sent by the 1st to the st duty officers in the past three months, recorded as sn （1） ~sn （st） ; Process A3: Get the daily working time, recorded as wh; define calculation formula A-1: ; Among them, sn （i） ,dd （i） andve （i） , respectively represent the number of telegrams sent by the i-th duty officer (in the past three months), the number of days on duty and the (average) telegram sending rate; the value range of i is: 1~st; (According to formula A-1) calculate the (average) telegram sending rate of the 1st to the st-th duty personnel, and get ve （1） ~ve （st） ; Process A4: Re （1） ~re （td） and （1） ~se （td） , respectively conduct time series analysis and construct the telegram receiving amount prediction equation and the telegram sending amount prediction equation; It should be noted that the unit of the time variable in "Process A4" of the present invention is "day", and the unit of the time period is also "day"; Process A41: The telegram reception amount prediction equation is used as equation A; according to re （1） ~re （td） Construct equation A; Process A411: （1） ~re （td）} as the time series Yt; use ADF detection to perform difference detection on Yt and obtain the difference order d; Taking d-order differences on Yt, we get ▽ （d） Yt; processed using autocorrelation function (ACF) and partial autocorrelation function (PACF) ▽ （d） Yt, get the autoregressive parameter p and sliding average parameter q of Yt; The value range of p and q is: [1, td) and the values ​​of (td-p) and (td-q) are both greater than or equal to 1; Process A412: Calculate re （1） ~re （td） The average values ​​are; Define calculation formula A-2: ;re （x） and ε （x） , respectively represent the amount of telegrams received on the xth day (in the past three months at the telegraph station) and the white noise (i.e., residual), and the value range of x is: 1~td; (According to formula A-2-1) calculate the white noise from the 1st to the tdth day: ε （1） ~ε （td） ; Calculate ε （1） ~ε （td） The variance of （1） ~ε （td）} as a white noise sequence εt; Process A413: Construct an initial expression of equation A as formula A-3: ; Where, k and l both represent time difference; B （k） Indicates the lag operator of Yt when the time difference is k, and the value range of k is: 1~p; BB （l） It represents the lag operator of εt when the time difference is l, and the value range of l is: 1~q; φ （k） Indicates B （k） The autocorrelation coefficient, θ （l） Indicates BB （l） The sliding average coefficient of , σ represents the error term; Process A414: Reset the amount of telegrams received on day x （x） Substituting into formula A-3 and expanding it, we get formula A-4: ; in, Represents （1） To（x） The d-order difference of Represents （1） To （x-1） The d-order difference of Represents （1） To （x-2） The d-order difference of ; and so on, Represents （1） To （x-p） The d-order difference of （x） with re （x-1） ~re （x-p） The extreme left value is re （1） , the extreme right value is re （td） ; (▽ （d） represents d-order difference operation) φ （1） ~φ （p） represents the 1st to pth order autocorrelation coefficient; ε （x） represents the white noise of the xth day, ε （x-1） represents the white noise of the (x-1)th day, ε （x-2） represents the white noise of the (x-2)th day; and so on, ε （x-q） represents the white noise of the (x-q)th day; ε （x） With ε （x-1） ~ε （x-q） The extreme left value is ε （1） , the extreme right value is ε （td） ; The value range of x is: 1~td; θ （1） ~θ （q） represents the 1st to qth order sliding mean coefficient; φ （1） ~φ （p） and θ （1） ~θ （q） The initial value of is 1; Process A415: Extract the left part of formula A-4 and calculate φ （1） ~φ （p） The specific value of The left side of formula A-4 is: ; Process A4151: First-order difference (i.e., let d = 1); for re （1） ~re （td） Perform first-order difference to obtain the difference value ri （1） ~ri （tf） ; Where tf represents the number of samples after the first-order difference, tf = td-1; Calculate ri （1） ~ri （tf）The average value of arc; repeated ε （1） ~ε （td） The calculation process is to calculate ri according to arc （1） ~ri （tf） The corresponding white noise is denoted by cε （1） ~cε （tf） ; Process A4152: In rc （1+p） ~rc （tf） Take any p consecutive subscript values ​​in as sample values, denoted as rc （sa） ~rc （sa+p） ; Among them, the value range of sa is: (1+p)~(tf-p); rc （sa） The extreme left value is ri （1+p） , rc （sa+p） The extreme right value is ri （tf） ; {rc （sa） ~rc （sa+p）} as set A; calculate the sample value of set A lagged by 1 day (i.e., rc （sa-1） ~rc （sa+p-1） ), and get the set A （1） ; The sample value lagged by 2 days (i.e., rc （sa-2） ~rc （sa+p-2） ), and get the set A （2） ; Similarly, the sample value lagged p days (i.e., rc （1） ~rc （sa） ), and get the set A （p） ; Process A4153: Calculate rc （sa） ~rc （sa+p） The variance of is denoted as σ(rc); According to set A and set A （1） ~Set A （p） , respectively calculate the sample value (ie rc （sa） ~rc （sa+p） ) The covariance of the lags 1 to p days, γ(rc （1） )~γ(rc （p） ); Define calculation formula A-5-1: rr （k） =γ(rc （k） ) / σ(rc);γ(rc （k） ) and rr （k） , respectively represent the covariance and correlation coefficient of the sample value lagged k days; (According to formula A-5-1) Calculate the correlation coefficient rr of the sample value lagged 1 to p days （1） ~rr（p） ; Process A4154: φ （1） ~φ （p） The first-order difference estimate of （1，1） ~φ （1，p） ; Construct an empty matrix of (p×1) and fill it with φ （1，1） ~φ （1，p） and rr （1） ~rr （p） , we get matrix AA and matrix AB; The matrix AA is: ; Matrix AB is: ; Among them, φ （1，2） Represents φ （2） The first-order difference estimate of , rr （2） Correlation coefficient of sample values ​​lagged 2 days; Process A4155: Determine whether (2×p-2)≤tf holds and define the calculation formula of AA; If not, the calculation formula of AA is: AA = (AB × AB T ) -1 ×AB; where × represents matrix multiplication, T represents the transpose of the matrix, and -1 represents the inverse of the matrix; If it is established, repeat process A4152 to process A4153 to calculate the correlation coefficient rr of the sample value lagged (-1) to [-(p-2)] days (i.e., 1 to (p-2) days in advance) （p+1） ~rr （2×p-2） ; Construct an empty matrix of (p×p) and fill it with σ(rc) and rr respectively. （1） ~rr （2×p-2） , we get the matrix AC; ; Among them, rr （2） ,rr （3） ,rr （4） and rr （p-1） , respectively represent the correlation coefficients of sample values ​​lagged by 2 days, 3 days, 4 days, and (p-1) days; The calculation formula of AA is: AA = (AC) -1 ×AB; Process A4156: Calculate φ (according to the calculation formula of AA defined in process A4155) （1，1） ~φ （1，p） The value of （1） ~qφ （p） ; Let the yth difference value be ri （y） ,ri （y） The corresponding white noise is cε（y） ; The value range of y is: (p+1)~tf; Construction formula A-5-2: ; where φ （1，k） represents the first-order difference estimate of the k-order autocorrelation coefficient, qφ （k） Represents φ （1，k） The specific value of ri （y-k） Represents （y） The difference value lagged k days; Represents （y） The estimated value of; the value range of k is: 1~p; Will ri （p+1） ~ri （tf） Substitute into formula A-5-2 to test and determine ri （p+1） ~ri （tf） The corresponding estimated value is （p+1） ~ri （tf） Is the difference within ξ? If yes, keep φ （1，1） ~φ （1，p） ; If not, modify φ （1，1） ~φ （1，p） Repeat process A4151 to process A4156 until it passes the test of formula A-5-2; Where ξ represents the error determination coefficient; (the value of ξ is 0.01; users or relevant technical personnel can adjust the value of ξ according to actual needs) Process A4157: Repeated calculation of qφ （1） ~qφ （p） The process of calculating φ （1） ~φ （p） The 1st to pth order autocorrelation coefficient in the second order difference: φ （2，1） ~φ （2，p） ; The 1st to pth order autocorrelation coefficients in the third-order difference: φ （3，1） ~φ （3，p） ; Similarly, the 1st to pth order autocorrelation coefficients of the d-order difference are: φ （d，1） ~φ （d，p） ; Calculate qφ （1） and φ （2，1） ~φ （d，1） The average value of φ （1） The value of Calculate qφ （2） and φ （2，2） ~φ （d，2） The average value of φ （2） The value of Similarly, calculate qφ （p） and φ （d，p） ~φ （d，p） The average value of φ （p） The value of Process A416: Extract the right part of formula A-4 and repeat the calculation of φ （1） ~φ （p） The same process as the specific value is used to calculate θ （1） ~θ （q） The specific value of The right side of formula A-4 is: ; φ （1） ~φ （p） The specific value of θ （1） ~θ （q） Substitute the specific value of into formula A-4 to obtain the telegram reception prediction equation, namely equation A; Process A42: Rewrite the process A41 (i.e., the process A411 to process A416) （1） ~re （td） Replace with se （1） ~se （td） , repeat the same process of constructing equation A to construct the telegram sending volume prediction equation; Process A5: Obtain the differential order d, and obtain the amount of telegrams received and sent by (telegraph station) on the previous d days (i.e., mo month (da-d-1) day and mo month (da-1) day), denoted as rv （1） ~rv （d） and sd （1） ~sd （d） ; rv （1） ~rv （d） and sd （1） ~sd （d） , respectively substituted into the telegram reception prediction equation and the telegram transmission prediction equation, calculate the total telegram reception and total telegram transmission of a single day (telegraph station), denoted as zrv and zse; Process A6: record the number of on-duty personnel in a single shift as pp, the number of days on duty as wd, and the maximum number of shifts on duty by a single person as ww; Definition of the mathematical model of the duty officer A-6-1: ; Among them, pe （i,j） It means that the i-th duty officer is on duty on the j-th day (in a week); Get the current working day, recorded as we (i.e. a day between Monday and Sunday); Set the value of j to we and set pe（i,j） The number of telegrams sent by the corresponding class staff (in the past three months), denoted as ve （i,j） ; (where ve （i,j） ∈{ve （1） ~ve （st）}) Define constraint A-6-2: ; Define the objective function A-6-3: ; Where ZZ represents the cost; Process A7: According to A-6-2 and A-6-3, use a linear programming solver (such as the simplex method) to solve the mathematical model A-6-1 and obtain the linear combination with the minimum cost as the initial duty sequence; Count the number of people in the initial duty sequence, recorded as pr; obtain the number of telegrams sent by the 1st to prth shift personnel (in the past three months), recorded as vee （1） ~vee （pr） ; (where {vee （1） ~vee （pr）}∈{ve （1） ~ve （st）}) Calculate vee （1） ~vee （pr） The sum of is denoted as ave; Get the amount of telegrams received by (telegraph station) yesterday, recorded as hrv; get the amount of telegrams received by (telegraph station) on the previous (d+1) day, recorded as rv （d+1） ; rv （2） ~rv （d+1） Substitute into the telegram sending volume prediction equation to calculate the expected telegram receiving volume of (telegraph station) yesterday, recorded as qrv; where rv （2） It indicates the number of telegrams received by (telegraph station) in the previous two days; Process A8: Determine whether hrv≤qrv is true; If established, the initial duty sequence remains unchanged; If it does not hold, change constraint A-6-2 to constraint A-6-4; Constraint A-6-4: ; Replace A-6-2 with A-6-4 and re-execute process A7 to obtain the optimal duty sequence; Process A9: Arrange the duty personnel according to the duty order of the optimal duty sequence and send the telegrams to be sent.

[0021] The telegram processing module includes: a telegram storage submodule and a telegram analysis submodule: Telegram storage submodule: used to parse and extract the telegram header and telegram tail of the quasi-telegram, obtain the receiving information and confidentiality level of the quasi-telegram; perform secondary encryption on the quasi-telegram according to the confidentiality level, and establish a storage structure according to the receiving information of the quasi-telegram to save the quasi-telegram; See also Figure 3 , Figure 3 The n in the structure represents the index of the leaf node, and the storage structure can be a structure for storing data such as a database, a blockchain, or a Merkle tree. When constructing the storage structure, the user or relevant technical personnel use "quasi-telegram received information" as the primary index and "quasi-telegram confidentiality level" as the secondary index, and set the same storage space for telegrams with the same received information and the same confidentiality level, and set different storage spaces for telegrams with the same received information and different confidentiality levels. Telegraph analysis submodule: used to analyze the changes in the electric signal of the quasi-telegraph and construct the telegraph equation; according to the telegraph equation, send the electric signal enhancement instruction to the lower-level telegraph station until the quasi-telegraph is received by the target unit; ("target unit" means the receiving unit of the quasi-telegraph) Because the telegram is a high-frequency electrical signal, energy loss will occur during the long-distance wireless transmission process. If the lower-level telegraph station does not perform electrical signal enhancement processing when receiving the "quasi-telegram" and sending it to the target unit, the information transmitted by the "quasi-telegram" will be distorted. Therefore, the telegraph station needs to send an "electrical signal enhancement instruction" to the lower-level telegraph station when transmitting the "quasi-telegram".

[0022] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technical personnel in this field according to actual conditions. For example, if there are weight coefficients and proportional coefficients, their set sizes are to quantify each parameter to obtain a specific value, which is convenient for subsequent comparison. Regarding the size of the weight coefficient and the proportional coefficient, it is sufficient as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0023] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A telegram management and analysis system based on big data and artificial intelligence, characterized in that: The system comprises: Telegram acquisition module: used to obtain received telegrams and obtain telegrams to be processed; Business management module: used to demodulate and back up the telegram to be processed to obtain the quasi-telegram; encrypt and modulate the quasi-telegram to obtain the telegram to be sent; Get the number of staff on duty, the number of telegrams sent by each staff on duty, and get the staff information; Obtain the amount of telegrams received and sent in the past three months to obtain historical telegram data; perform time series analysis on historical telegram data to calculate the expected receiving rate and expected sending rate of telegram sending work; obtain duty requirements, and schedule and adjust the duty personnel in combination with employee information, and send the telegrams to be sent; The telegram processing module includes: a telegram storage submodule and a telegram analysis submodule: Telegram storage submodule: used to parse and extract the telegram header and telegram tail of the quasi-telegram, obtain the receiving information and confidentiality level of the quasi-telegram; perform secondary encryption on the quasi-telegram according to the confidentiality level, and establish a storage structure according to the receiving information of the quasi-telegram to save the quasi-telegram; Telegraph analysis submodule: used to analyze the changes in the electrical signals of the quasi-telegraph and construct telegraph equations; according to the telegraph equations, electrical signal enhancement instructions are sent to the lower-level telegraph stations until the quasi-telegraph is received by the target unit.

2. According to the big data and artificial intelligence telegram management and analysis system of claim 1, it is characterized in that: The workflow of the business management module is as follows: Process A1: Duty requirements include: daily working hours, duty days, number of staff on duty in a single shift, and maximum number of shifts for a single person on duty; Count the number of telegrams to be processed te; demodulate and back up the telegrams to be processed to obtain quasi-telegrams; encrypt and modulate the quasi-telegrams to obtain telegrams to be sent; Process A2: Get the number of on-duty personnel st; get the number of telegrams received in the past three months re （1） ~re （td） ; Telegram sending volume （1） ~se （td） ; Get the number of days dd that the 1st to the st duty personnel have been on duty （1） ~dd （st） , the amount of telegrams sent sn （1） ~sn （st） ; Process A3: Get the daily working time wh; define calculation formula A-1: ; Among them, sn （i） ,dd （i） andve （i） , respectively represent the number of telegrams sent by the i-th duty officer, the number of days on duty, and the rate of telegrams sent; Calculate the telegram sending rate of the 1st to the st-th duty personnel, and get ve （1） ~ve （st） ; Process A4: Re （1） ~re （td） and （1） ~se （td） , respectively conduct time series analysis and construct the telegram receiving amount prediction equation and the telegram sending amount prediction equation; Process A5: Get the differential order d, and get the telegram reception volume rv in the previous d days （1） ~rv （d） , telegram sending volume sd （1） ~sd （d） ; rv （1） ~rv （d） and sd （1） ~sd （d） , substitute them into the telegram reception prediction equation and the telegram sending prediction equation respectively, and calculate the total telegram reception volume zrv and the total telegram sending volume zse for a single day.

3. According to the big data and artificial intelligence telegram management and analysis system of claim 2, it is characterized in that: The subsequent process of process A5 is as follows: Process A6: record the number of on-duty personnel in a single shift as pp, the number of days on duty as wd, and the maximum number of shifts on duty by a single person as ww; Definition of the mathematical model of the duty officer A-6-1: ； Among them, pe （i,j） It means that the i-th duty officer is on duty on the j-th day; Get the current working day we; set the value of j to we and set pe （i,j） The corresponding number of telegrams sent by the class personnel is denoted as ve （i,j） ; Define constraint A-6-2: ; Define the objective function A-6-3: ; Where ZZ represents the cost; Process A7: According to A-6-2 and A-6-3, use the linear programming solver to solve the mathematical model A-6-1 and obtain the linear combination with the minimum cost as the initial duty sequence; Get yesterday's telegram reception volume hrv; the previous (d+1) day's telegram reception volume rv （d+1） ; rv （2） ~rv （d+1） Substitute into the telegram sending volume prediction equation to calculate yesterday's expected telegram receiving volume qrv; where rv （2） It indicates the amount of telegrams received in the previous 2 days; Process A8: Determine whether hrv≤qrv is true; If established, the initial duty sequence remains unchanged; If it does not hold, change constraint A-6-2 to constraint A-6-4: ； Replace A-6-2 with A-6-4 and re-execute process A7 to obtain the optimal duty sequence; Process A9: Arrange the duty personnel according to the duty order of the optimal duty sequence and send the telegrams to be sent.

4. The big data and artificial intelligence telegram management and analysis system according to claim 2 is characterized in that: The specific process of process A4 is as follows: Process A41: The telegram reception amount prediction equation is used as equation A; according to re （1） ~re （td） Construct equation A; Process A42: Re （1） ~re （td） Replace with se （1） ~se （td） , repeat the same process of constructing equation A to construct the telegram sending volume prediction equation.

5. The big data and artificial intelligence telegram management and analysis system according to claim 4 is characterized in that: The specific process of process A41 is as follows: Process A411: （1） ~re （td） } as the time series Yt; use ADF detection to perform difference detection on Yt and obtain the difference order d; Taking d-order differences on Yt, we get ▽ （d） Yt; processed using autocorrelation function and partial autocorrelation function▽ （d） Yt, get the autoregressive parameter p and sliding average parameter q of Yt; Process A412: Calculate re （1） ~re （td） The average values ​​are; Define calculation formula A-2: ;re （x） and ε （x） , respectively represent the amount of telegrams received and white noise on the xth day; Calculate the white noise from day 1 to day td: ε （1） ~ε （td） ; Set {ε （1） ~ε （td） } as a white noise sequence εt; Process A413: Construct an initial expression of equation A as formula A-3: ； Where, k and l both represent time difference; B （k） represents the lag operator of Yt when the time difference is k; BB （l） represents the lag operator of εt when the time difference is l; φ （k） Indicates B （k） The autocorrelation coefficient, θ （l） Indicates BB （l） is the sliding average coefficient of , and σ represents the error term.

6. A telegram management and analysis system of big data and artificial intelligence according to claim 5, characterized in that: The subsequent process of the process A413 is as follows: Process A414: Reset the amount of telegrams received on day x （x） Substituting into formula A-3 and expanding it, we get formula A-4: ； in, Represents （1） To （x） The d-order difference of Represents （1） To （x-1） The d-order difference of Represents （1） To （x-p） The d-order difference of （1） ~φ （p） represents the 1st to pth order autocorrelation coefficient; ε （x） represents the white noise of the xth day, ε （x-1） represents the white noise of the (x-1)th day; similarly, ε （x-q） represents the white noise of the (x-q)th day; θ （1） ~θ （q） represents the 1st to qth order sliding mean coefficient; Process A415: Extract the left part of formula A-4 and calculate φ （1） ~φ （p） The specific value of The left side of formula A-4 is: ； Process A416: Extract the right part of formula A-4 and repeat the calculation of φ （1） ~φ （p） The same process as the specific value is used to calculate θ （1） ~θ （q） The specific value of The right side of formula A-4 is: ； φ （1） ~φ （p） The specific value of θ （1） ~θ （q） Substitute the specific value of into formula A-4 to obtain the telegram reception prediction equation.

7. The big data and artificial intelligence telegram management and analysis system according to claim 6 is characterized in that: The specific process of process A415 is as follows: Process A4151: First-order difference; for re （1） ~re （td） Perform first-order difference to obtain the difference value ri （1） ~ri （tf） ; Where tf represents the number of samples after the first-order difference, tf = td-1; Calculate ri （1） ~ri （tf） The corresponding white noise cε （1） ~cε （tf） ; Process A4152: In rc （1+p） ~rc （tf） Take any p consecutive subscript values ​​as sample values, denoted as rc （sa） ~rc （sa+p） ; {rc （sa） ~rc （sa+p） } as set A; calculate the sample value of set A lagged by 1 day, and get set A （1） Similarly, calculate the sample value of the lag p days and get the set A （p） ; Process A4153: Calculate rc （sa） ~rc （sa+p） The variance of is denoted as σ(rc); According to set A and set A （1） ~Set A （p） , calculate the covariance of the sample values ​​lagged 1 to p days respectively, and get γ(rc （1） )~γ(rc （p） ); Define calculation formula A-5-1: rr （k） =γ(rc （k） ) / σ(rc);γ(rc （k） ) and rr （k） , respectively represent the covariance and correlation coefficient of the sample value lagged k days; Calculate the correlation coefficient rr of the sample value lagged 1 to p days （1） ~rr （p） .

8. The big data and artificial intelligence telegram management and analysis system according to claim 7, characterized in that: The subsequent process of process A4153 is as follows: Process A4154: φ （1） ~φ （p） The first-order difference estimate of （1，1） ~φ （1，p） ; Construct an empty matrix of (p×1) and fill it with φ （1，1） ~φ （1，p） and rr （1） ~rr （p） , we get matrix AA and matrix AB; The matrix AA is: ; Matrix AB is: ; Among them, φ （1，2） Represents φ （2） The first-order difference estimate of , rr （2） Correlation coefficient of sample values ​​lagged 2 days; Process A4155: Determine whether (2×p-2)≤tf holds and define the calculation formula of AA; If not, the calculation formula of AA is: AA = (AB × AB T ) -1 ×AB; If it is established, repeat process A4152 to process A4153 to calculate the correlation coefficient rr of the sample value lagged (-1) to [-(p-2)] days （p+1） ~rr （2×p-2） ; Construct an empty matrix of (p×p) and fill it with σ(rc) and rr respectively. （1） ~rr （2×p-2） , we get the matrix AC; ; Among them, rr （2） ,rr （3） ,rr （4） and rr （p-1） , respectively represent the correlation coefficients of sample values ​​lagged by 2 days, 3 days, 4 days, and (p-1) days; The calculation formula of AA is: AA = (AC) -1 ×AB.

9. The big data and artificial intelligence telegram management and analysis system according to claim 8, characterized in that: The subsequent process of process A4155 is as follows: Process A4156: Calculate φ （1，1） ~φ （1，p） The value of （1） ~qφ （p） ; Let the yth difference value be ri （y） ,ri （y） The corresponding white noise is cε （y） ; Construction formula A-5-2: ; where qφ （k） Represents the specific value of the first-order difference estimate of the k-order autocorrelation coefficient; ri （y-k） Represents （y） The difference value lagged k days; Represents （y） An estimated value of will ri （p+1） ~ri （tf） Substitute into formula A-5-2 to test and determine ri （p+1） ~ri （tf） The corresponding estimated value is （p+1） ~ri （tf） Whether the difference is within ξ; If yes, keep φ （1，1） ~φ （1，p） ; If not, modify φ （1，1） ~φ （1，p） The value of is taken, and the process A4151 to process A4156 are repeatedly executed until the test of formula A-5-2 is passed; wherein ξ represents the error determination coefficient; Process A4157: Repeated calculation of qφ （1） ~qφ （p） The process of calculating φ （1） ~φ （p） The 1st to pth order autocorrelation coefficient in the second-order difference: φ （2，1） ~φ （2，p） ; Similarly, the 1st to pth order autocorrelation coefficient of the d-order difference is: φ （d，1） ~φ （d，p） ; Calculate qφ （1） and φ （2，1） ~φ （d，1） The average value of φ （1） The value of Similarly, calculate qφ （p） and φ （d，p） ~φ （d，p） The average value of φ （p） The value of .

Images