Personnel falling-into-water comprehensive detection method based on 5G

Abstract

The invention discloses a personnel falling-into-water comprehensive detection method based on 5G, and belongs to the technical field of water search and rescue, in the method, a person carries a 5G terminal, and the 5G terminal detects the altitude where the person is located in real time and sends personnel falling-into-water information A to a falling-into-water comprehensive detection server when a set altitude judgment condition is met; the 5G base station detects an uplink signal sent by the 5G terminal in real time, and sends personnel falling-into-water information B to the falling-into-water comprehensive detection server when a set uplink signal judgment condition is met; the 5G terminal detects a downlink signal issued by the 5G base station in real time, and sends personnel falling-into-water information C to the falling-into-water comprehensive detection server when a set downlink signal judgment condition is met; and the falling-into-water comprehensive detection server comprehensively judges that the personnel fall into water according to the personnel falling-into-water information A, the personnel falling-into-water information B and the personnel falling-into-water information C. The technical problems that a traditional falling-into-water detection method is single in falling-into-water detection means, and false alarm and missing alarm are prone to occurring in a complex scene are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of water search and rescue, and specifically relates to a 5G-based comprehensive method for detecting people falling into water. Background Art

[0002] In terms of falling-into-water detection, taking ship application scenarios as an example, the most commonly used ones are the ORCA falling-into-water alarm system and the LIFETAG personnel falling-into-water wireless alarm system.

[0003] The ORCA drowning alarm system is mainly divided into four parts: the distress terminal, the shipboard terminal, the wireless terminal and the tracking module. Five seconds after a person falls into the water, the distress terminal will automatically send a distress message, and at the same time send the personal information and location information of the drowning person to the shipboard terminal. After receiving the distress message, the shipboard terminal will automatically alarm and display the location information of the drowning person on the screen. The shipboard terminal can also send the location information of the drowning person to the tracker for monitoring and rescue organization.

[0004] LIFE TAG's wireless alarm system for people falling into water mainly consists of a base station, a pager and an alarm. When the person carrying the pager exceeds the transmission distance of the base station signal, the base station cannot connect to the pager and the base station will automatically sound the whistle to alarm.

[0005] However, the ORCA water fall alarm system and the LIFE TAG wireless water fall alarm system have a single water fall detection method, and are prone to false alarms and missed alarms when facing complex scenarios. In addition, the base station of the LIFE TAG wireless water fall alarm system only supports a maximum of 16 life-saving trigger devices. If the number exceeds 16, multiple base stations need to be expanded through the system, which is costly. Summary of the invention

[0006] In view of this, the present invention provides a 5G-based comprehensive detection method for people falling into water. The comprehensive detection method integrates multiple detection means to solve the technical problems of traditional fall-in-water detection methods, such as single fall-in-water detection means, prone to false alarms and missed alarms in complex scenarios, and a limited number of supported life-saving trigger devices.

[0007] The 5G-based comprehensive method for detecting people falling into water adopts the following technical solutions:

[0008] The personnel carry a 5G terminal, and: the 5G terminal detects the personnel's altitude in real time, and sends personnel-falling-into-water information A to the comprehensive water-falling detection server when the set altitude judgment condition is met; the 5G base station detects the uplink signal sent by the 5G terminal in real time, and sends personnel-falling-into-water information B to the comprehensive water-falling detection server when the set uplink signal judgment condition is met; the 5G terminal detects the downlink signal sent by the 5G base station in real time, and sends personnel-falling-into-water information C to the comprehensive water-falling detection server when the set downlink signal judgment condition is met; the comprehensive water-falling detection server comprehensively determines whether the personnel has fallen into the water based on personnel-falling-into-water information A, personnel-falling-into-water information B and personnel-falling-into-water information C.

[0009] Furthermore, the altitude determination condition is:

[0010] During the set time period A, the altitude keeps decreasing and the average decrease is greater than the set threshold altitude.

[0011] Furthermore, the uplink signal determination condition is:

[0012] Within the set time length B, the average received power of the uplink signal is less than the set uplink signal threshold power, and the set time length B is greater than the set time length A.

[0013] Furthermore, the downlink signal determination condition is:

[0014] Within the set duration C, the SSB average received power is less than the set SSB threshold power and the set duration C is greater than the set duration A.

[0015] Furthermore, the set duration B is equal to the set duration C.

[0016] Furthermore, the credibility weight of the person-overboard information A is recorded as W1, the credibility weight of the person-overboard information B is recorded as W2, the credibility weight of the person-overboard information C is recorded as W3, and the total credibility weight is recorded as W;

[0017] When W is greater than the set threshold credibility, the comprehensive water-falling detection server determines that a person has fallen into the water, and W is calculated according to the following formula:

[0018] W=N1*W1+N2*W2+N3*W3

[0019] in:

[0020] “*” is a multiplication sign;

[0021] N1 takes a value of 0 or 1. When the integrated water-falling detection server receives the information A of a person falling into the water, N1 takes a value of 1, otherwise it takes a value of 0;

[0022] N2 takes a value of 0 or 1. When the integrated water-falling detection server receives the information B of a person falling into the water, N2 takes a value of 1, otherwise it takes a value of 0;

[0023] N3 takes a value of 0 or 1. When the integrated water-falling detection server receives the information C of a person falling into the water, N3 takes a value of 1, otherwise it takes a value of 0;

[0024] W1+W2+W3=1, and W1, W2, and W3 are all smaller than the threshold credibility.

[0025] Furthermore, the sum of W1 and W2 is greater than the threshold credibility, the sum of W1 and W3 is less than the threshold credibility, and the sum of W2 and W3 is less than the threshold credibility.

[0026] Furthermore, W2 is equal to W3.

[0027] The present invention also provides another method for comprehensively determining whether a person has fallen into water based on the person-falling-into-water information A, the person-falling-into-water information B and the person-falling-into-water information C, wherein:

[0028] When the actual probability of falling into water P is greater than the preset threshold value, the comprehensive detection server determines that a person has fallen into water. P is calculated according to the following formula:

[0029] P=N1*P(1)+N2*P(2)+N3*P(3)

[0030] “*” is a multiplication sign;

[0031] The value of N1 is 0 or 1. When the comprehensive water-falling detection server receives the information A of a person falling into the water, the value of N1 is 1, otherwise the value is 0;

[0032] N2 takes a value of 0 or 1. When the comprehensive water-falling detection server receives the information B of a person falling into the water, N2 takes a value of 1, otherwise it takes a value of 0;

[0033] N3 takes a value of 0 or 1. When the integrated water-falling detection server receives the information C of a person falling into the water, N3 takes a value of 1, otherwise it takes a value of 0;

[0034] P(1) is the probability of a person actually falling into the water when the comprehensive detection server uses Bayes’ theorem to infer the person falling into the water information A;

[0035] P(2) is the probability of a person actually falling into the water when the comprehensive detection server uses Bayes’ theorem to infer the person falling into the water information B;

[0036] P(3) is the probability of a person actually falling into the water when the comprehensive detection server uses Bayes' theorem to infer the probability of a person actually falling into the water when receiving the person falling into the water information C.

[0037] Furthermore, before the comprehensive water-fall detection server uses Bayesian theorem to infer P(1), P(2) and P(3), it includes:

[0038] Obtain prior probability: obtain the prior probability "P(falling into water)" of a person falling into water in the falling-into-water detection area, the probability "P(receiving person falling-into-water information A)" of the comprehensive falling-into-water detection server receiving person falling-into-water information B "P(receiving person falling-into-water information B)", and the probability "P(receiving person falling-into-water information C)" of the comprehensive falling-into-water detection server receiving person falling-into-water information C;

[0039] Collect observation data: collect the probability that the comprehensive detection server for falling into water receives the information A of a person falling into water when a falling-incident occurs, "P(received information about a person falling into water|falling into water)", the probability that the comprehensive detection server for falling into water receives the information B of a person falling into water when a falling-incident occurs, "P(received information about a person falling into water B|falling into water)", and the probability that the comprehensive detection server for falling into water receives the information C of a person falling into water when a falling-incident occurs, "P(received information about a person falling into water C|falling into water)".

[0040] Beneficial effects:

[0041] 1. The comprehensive detection server for falling into water comprehensively determines the falling of people based on the falling-into-water information A, falling-into-water information B and falling-into-water information C, solving the technical problems of the traditional falling-into-water detection method, which has a single falling-into-water detection method, is prone to false alarms and missed alarms in complex scenarios, and has a limited number of supported life-saving trigger devices. Moreover, a 5G base station can support thousands or even tens of thousands of 5G terminals, solving the technical problem of the limited number of trigger devices supported by the base station of the traditional LIFE TAG wireless alarm system for falling-into-water people.

[0042] 2. The altitude determination condition is set as follows: within the set time length A, the altitude has been decreasing and the average decrease is greater than the set threshold altitude. In this way, it can avoid the problem of false alarms of falling into water when going downstairs quickly or falling, improve the accuracy of people falling into water detection, and reduce the false alarm and missed alarm rates.

[0043] 3. The uplink signal determination condition is set as follows: within the set duration B, the average uplink signal received power is less than the set uplink signal threshold power, and the set duration is greater than the set duration A; the downlink signal determination condition is set as follows: within the set duration C, the SSB average received power is less than the set SSB threshold power, and the set duration B is greater than the set duration A.

[0044] In this way, it is possible to avoid the problem of false alarms of people falling into the water due to the influence of 5G signal strength when people pass by metal structures, thereby improving the accuracy of people falling into the water detection and reducing the false alarm and missed alarm rates.

[0045] 4. The sum of W1 and W2 is greater than the threshold credibility, the sum of W1 and W3 is less than the threshold credibility, and the sum of W2 and W3 is less than the threshold credibility. Because the altitude condition is more relevant to the situation of people falling into the water than the uplink signal judgment condition and the downlink signal judgment condition, at least the altitude judgment condition must be met before it can be judged that a person has fallen into the water, which can improve the accuracy of people falling into the water detection and reduce the false alarm and missed alarm rates.

[0046] 5. The credibility weight of the person-falling-into-water information A is recorded as W1, the credibility weight of the person-falling-into-water information B is recorded as W2, the credibility weight of the person-falling-into-water information C is recorded as W3, and the total credibility weight is recorded as W; when W is greater than the set threshold credibility, the comprehensive water-falling detection server determines that a person has fallen into the water.

[0047] In this way, we can first conduct an experiment, and then reasonably determine the weight of the corresponding person-overboard information based on the false alarm rates of the three detection methods in the experiment, thereby improving the accuracy of person-overboard detection and reducing the false alarm and missed alarm rates.

[0048] 6. Because the coupling degree between the uplink signal and the downlink signal is high, making the credibility weight of the person-falling-into-water information B equal to the credibility weight of the person-falling-into-water information C can match the actual situation, thereby improving the accuracy of person-falling-into-water detection and reducing the false alarm and missed alarm rates.

[0049] 7. Based on the comprehensive use of altitude information, uplink signals, and downlink signals, the Bayesian theorem is used to determine whether a person has fallen into the water. This can continuously update the prior probability and continuously improve the accuracy of detecting when a person has fallen into the water. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of the comprehensive detection method of people falling into water based on 5G of the present invention;

[0051] Figure 2 for Figure 1 Schematic diagram of the process of detecting people falling into water based on 5G terminal sensors;

[0052] Figure 3 for Figure 1 Schematic diagram of the process of detecting people falling into water based on uplink signal analysis;

[0053] Figure 4 for Figure 1 Schematic diagram of the process of detecting people falling into water based on downlink signal analysis. DETAILED DESCRIPTION

[0054] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0055] Embodiment 1:

[0056] like Figure 1-Figure 4As shown in the figure, a 5G-based comprehensive detection method for people falling into water is provided. The method requires people to carry 5G terminals, which includes three detection methods, namely:

[0057] (I) Detection of people falling into water based on 5G terminal sensors:

[0058] Commercial 5G terminals generally integrate barometers (or other sensors) to measure altitude information, so the altitude information of commercial 5G terminals is available. Figure 2 As shown, the 5G terminal monitors the altitude of the personnel in real time, and sends the personnel falling into water information A to the comprehensive falling into water detection server through the 5G base station when the monitored altitude reaches the set altitude judgment condition;

[0059] (II) Detection of people falling into water based on uplink signal analysis:

[0060] Uplink signal refers to the signal sent by commercial 5G terminals to 5G base stations; uplink signal analysis refers to the power detection of uplink signals received by 5G base stations (which can be regarded as monitoring the uplink signal strength). According to the 5G network coverage design, the uplink signal receiving power should not be lower than the set threshold power, and the uplink signal judgment condition can be set accordingly. If the uplink signal judgment condition is met, the 5G base station generates a report of someone falling into the water and sends it to the comprehensive detection server for falling into the water. Figure 3 As shown, the 5G base station detects the power of the uplink signal sent by the received 5G terminal in real time, and sends the person falling into the water information B to the comprehensive detection server for falling into the water when the set uplink signal judgment condition is met;

[0061] (III) Detection of people falling into water based on downlink signal analysis

[0062] Downlink signal refers to the broadcast channel SSB sent by 5G base stations to commercial 5G terminals; downlink signal analysis refers to the power detection of the broadcast channel SSB sent by the 5G base station by the commercial 5G terminal (which can be regarded as monitoring the downlink signal strength). According to the 5G network coverage design, the receiving power of the downlink SSB should not be lower than the set threshold power, and the downlink signal judgment condition can be set accordingly. Figure 4 As shown, the 5G terminal detects the power of the broadcast channel SSB sent by the 5G base station in real time, and sends the person falling into the water information C to the comprehensive detection server for falling into the water when the set downlink signal judgment condition is met;

[0063] In this embodiment, the comprehensive water-falling detection server comprehensively determines whether a person has fallen into the water through the person's altitude information, the uplink signal received by the 5G base station, and the broadcast channel SSB received by the commercial 5G terminal, solving the technical problems of the traditional water-falling detection method, which has a single water-falling detection method, is prone to false alarms and missed alarms in complex scenarios, and has a limited number of supported life-saving trigger devices. Moreover, a 5G base station can support thousands or even tens of thousands of 5G terminals, solving the technical problem of the limited number of trigger devices supported by the base station of the traditional LIFE TAG wireless alarm system for people falling into the water.

[0064] The above-mentioned altitude determination condition can be set as the altitude has been decreasing within the set time length, the uplink signal determination condition can be set as the average uplink signal received power is less than the set uplink signal threshold power within the set time length, and the downlink signal determination condition can be set as the average received power is less than the set SSB threshold power within the set time length. It can be understood that the altitude condition, the uplink signal determination condition and the downlink signal determination condition can also be set in other reasonable ways. In this embodiment, the altitude determination condition is: within the set time length A, the altitude has been decreasing and the average decrease is greater than the set threshold altitude. In this way, it can avoid the problem of false alarm of falling into water when going downstairs quickly or falling to the greatest extent, improve the accuracy of personnel falling into water detection, and reduce the false alarm and missed alarm rate. The uplink signal determination condition is set as: within the set time length B, the average uplink signal received power is less than the set uplink signal threshold power, and the set time length B is greater than the set time length A; the downlink signal determination condition is set as: within the set time length C, the SSB average received power is less than the set SSB threshold power, and the set time length C is greater than the set time length A. In this way, the problem of false alarms of people falling into the water caused by the 5G signal strength being affected when people pass by metal structures can be avoided to the greatest extent, the accuracy of people falling into the water detection can be improved, and the false alarm and missed alarm rates can be reduced.

[0065] In this embodiment, considering that when a person falls into water, the comprehensive detection server for falling into water may receive information reported by multiple detection means, based on which an alarm credibility weight is set for each detection means, and only when the sum of the credibility weights is greater than the set threshold, a falling into water alarm is issued. In this way, the weight of the corresponding person falling into water information can be reasonably determined according to the false alarm rates of the three detection means in the experiment, thereby improving the accuracy of person falling into water detection and reducing the false alarm and missed alarm rates.

[0066] For the convenience of description, the credibility weight of the person-falling-into-water information A is recorded as W1, the credibility weight of the person-falling-into-water information B is recorded as W2, the credibility weight of the person-falling-into-water information C is recorded as W3, and the total credibility weight is recorded as W. In this embodiment, when W is greater than the set threshold credibility, the comprehensive detection server for falling into water determines that there is a person falling into water, and W is calculated according to the following formula:

[0067] W=N1*W1+N2*W2+N3*W3

[0068] in:

[0069] “*” is a multiplication sign;

[0070] N1 takes a value of 0 or 1. When the integrated water-falling detection server receives the information A of a person falling into the water, N1 takes a value of 1, otherwise it takes a value of 0;

[0071] N2 takes a value of 0 or 1. When the comprehensive water-falling detection server receives the information B of a person falling into the water, N2 takes a value of 1, otherwise it takes a value of 0;

[0072] N3 takes a value of 0 or 1. When the integrated water-falling detection server receives the information C of a person falling into the water, N3 takes a value of 1, otherwise it takes a value of 0;

[0073] W1+W2+W3=1, and W1, W2, and W3 are all less than the threshold credibility.

[0074] In this embodiment, upon receiving at least two of the person-falling-into-water information A, the person-falling-into-water information B and the person-falling-into-water information C, the comprehensive water-falling detection server determines that a person has fallen into the water.

[0075] Furthermore, at least two pieces of water-falling information include the person-falling-into-water information A, that is, at least the altitude condition must be met (that is, the sum of the credibility weights of W2 and W3 is less than the threshold credibility, and the sum of the credibility weights of W1 and W2 and the sum of the credibility weights of W1 and W3 are both greater than the threshold credibility). Because the altitude condition is more relevant to the situation of a person falling into water than the uplink signal determination condition and the downlink signal determination condition, therefore, setting the weight of the person-falling-into-water information A related to the altitude condition to the maximum can improve the accuracy of the person-falling-into-water detection and reduce the false alarm and missed alarm rates. Moreover, considering that the coupling degree between the uplink signal and the downlink signal is high, in this embodiment, the credibility weight W2 of the person-falling-into-water information B is equal to the credibility weight W3 of the person-falling-into-water information C, and the set duration B is equal to the set duration C, which can better match the actual situation, thereby improving the accuracy of the person-falling-into-water detection and reducing the false alarm and missed alarm rates.

[0076] Specifically, in this embodiment, the false alarm rates of the above three detection methods are experimentally analyzed. During the test, 500 water fall alarms reported by each detection method are collected and counted, and the number of false alarms is counted to obtain the false alarm rate value of each method. The test results of the three detection methods are shown in the following table:

[0077] Detection methods Total number of alarms False positives 5G uplink signal strength detection 500 108 5G downlink signal strength detection 500 99 Mobile device altitude detection 500 49

[0078] According to the test results and the above-mentioned weight setting requirements, the weight values ​​of the three proposed detection methods are shown in the following table:

[0079] Detection methods Test method number Credibility Weight 5G uplink signal strength detection No.1 0.25 5G downlink signal strength detection No.2 0.25 Mobile device altitude detection No.3 0.5

[0080] That is to say, the total credibility weight W = No.3*0.5+No.2*0.25+No.1*0.25, among which No.3 takes the value of 0 or 1. When the comprehensive detection server for falling into water receives the information A of a person falling into water, No.3 takes the value of 1, otherwise it takes the value of 0; No.2 takes the value of 0 or 1. When the comprehensive detection server for falling into water receives the information B of a person falling into water, No.2 takes the value of 1, otherwise it takes the value of 0; No.1 takes the value of 0 or 1. When the comprehensive detection server for falling into water receives the information C of a person falling into water, No.1 takes the value of 1, otherwise it takes the value of 0.

[0081] Embodiment 2:

[0082] This embodiment provides a comprehensive water-falling detection method using Bayesian theorem. Similar to the first embodiment, the comprehensive water-falling detection server still comprehensively determines whether a person falls into the water by using the altitude information of the person, the uplink signal information sent by the 5G terminal to the 5G base station, and the broadcast channel SSB information (downlink signal information) sent by the 5G terminal to the 5G base station. However, the specific comprehensive judgment strategy of this embodiment uses Bayesian theorem, specifically, including the following operations:

[0083] Obtaining prior probability: the prior probability of a person falling into water in the falling-into-water detection area "P(falling-into-water)", the probability of the falling-into-water integrated detection server receiving the person falling-into-water information A "P(receiving the person falling-into-water information A)", the probability of the falling-into-water integrated detection server receiving the person falling-into-water information B "P(receiving the person falling-into-water information B)", and the probability of the falling-into-water integrated detection server receiving the person falling-into-water information C "P(receiving the person falling-into-water information C)" can be obtained through historical data or experience;

[0084] Collect observation data: collect the probability that the comprehensive detection server for falling into water receives the information A of a person falling into water when the falling into water event occurs, "P(received information A of a person falling into water|falling into water)", the probability that the comprehensive detection server for falling into water receives the information B of a person falling into water when the falling into water event occurs, "P(received information B of a person falling into water|falling into water)", and the probability that the comprehensive detection server for falling into water receives the information C of a person falling into water when the falling into water event occurs, "P(received information C of a person falling into water|falling into water)".

[0085] Inference using Bayesian theorem: The comprehensive detection server for falling into water uses Bayesian theorem to infer the probability of a person falling into water when receiving information A of a person falling into water, "P(falling into water|receiving information A of a person falling into water)", calculated as "P(1)", the probability of a person falling into water when the comprehensive detection server for falling into water receives information B of a person falling into water, "P(falling into water|receiving information B of a person falling into water)", calculated as "P(2)", and the probability of a person falling into water when the comprehensive detection server for falling into water receives information C of a person falling into water, "P(falling into water|receiving information C of a person falling into water)", calculated as "P(3)";

[0086] When the actual probability of falling into water P is greater than the preset threshold value, the comprehensive detection server determines that a person has fallen into water. P is calculated according to the following formula:

[0087] P=N1*P(1)+N2*P(2)+N3*P(3)

[0088] N1 takes a value of 0 or 1. When the integrated water-falling detection server receives the information A of a person falling into the water, N1 takes a value of 1, otherwise it takes a value of 0;

[0089] N2 takes a value of 0 or 1. When the comprehensive water-falling detection server receives the information B of a person falling into the water, N2 takes a value of 1, otherwise it takes a value of 0;

[0090] N3 takes a value of 0 or 1. When the integrated water-falling detection server receives the information C of a person falling into the water, N3 takes a value of 1, otherwise it takes a value of 0;

[0091] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A 5G-based comprehensive method for detecting people falling into water, characterized in that: The personnel carry 5G terminals and: The 5G terminal detects the altitude of the person in real time, and sends the person falling into water information A to the comprehensive falling into water detection server when the set altitude judgment condition is met; The 5G base station detects the uplink signal sent by the 5G terminal in real time, and sends the person falling into the water information B to the comprehensive detection server when the set uplink signal judgment condition is met; The 5G terminal detects the downlink signal sent by the 5G base station in real time, and sends the person falling into the water information C to the comprehensive detection server when the set downlink signal judgment conditions are met; The comprehensive water-falling detection server comprehensively determines whether a person has fallen into the water based on the person-falling-into-water information A, the person-falling-into-water information B and the person-falling-into-water information C.

2. According to claim 1, a 5G-based comprehensive detection method for people falling into water is characterized in that: The altitude determination conditions are: During the set time period A, the altitude keeps decreasing and the average decrease is greater than the set threshold altitude.

3. According to claim 2, a 5G-based comprehensive detection method for people falling into water is characterized in that: The uplink signal determination condition is: Within the set time length B, the average received power of the uplink signal is less than the set uplink signal threshold power, and the set time length B is greater than the set time length A.

4. According to the 5G-based comprehensive detection method for people falling into water according to claim 3, it is characterized in that: The downlink signal determination condition is: Within the set duration C, the SSB average received power is less than the set SSB threshold power and the set duration C is greater than the set duration A.

5. According to claim 4, a 5G-based comprehensive detection method for people falling into water is characterized in that: The set duration B is equal to the set duration C.

6. A 5G-based comprehensive method for detecting people falling into water according to claim 5, characterized in that: The credibility weight of the person-overboard information A is recorded as W1, the credibility weight of the person-overboard information B is recorded as W2, the credibility weight of the person-overboard information C is recorded as W3, and the total credibility weight is recorded as W; When W is greater than the set threshold credibility, the comprehensive water-falling detection server determines that a person has fallen into the water, and W is calculated according to the following formula: W=N1*W1+N2*W2+N3*W3 in: "*" is a multiplication sign; N1 takes a value of 0 or 1. When the integrated water-falling detection server receives the information A of a person falling into the water, N1 takes a value of 1, otherwise it takes a value of 0; N2 takes a value of 0 or 1. When the comprehensive water-falling detection server receives the information B of a person falling into the water, N2 takes a value of 1, otherwise it takes a value of 0; N3 takes a value of 0 or 1. When the integrated water-falling detection server receives the information C of a person falling into the water, N3 takes a value of 1, otherwise it takes a value of 0; W1+W2+W3=1, and W1, W2, and W3 are all smaller than the threshold credibility.

7. A 5G-based comprehensive method for detecting people falling into water according to claim 6, characterized in that: The sum of W1 and W2 is greater than the threshold credibility, the sum of W1 and W3 is less than the threshold credibility, and the sum of W2 and W3 is less than the threshold credibility.

8. A 5G-based comprehensive method for detecting people falling into water according to claim 6 or 7, characterized in that: W2 is equal to W3.

9. The 5G-based comprehensive method for detecting people falling into water according to claim 1, characterized in that: When the actual probability of falling into water P is greater than the preset threshold value, the comprehensive detection server determines that a person has fallen into water. P is calculated according to the following formula: P=N1*P(1)+N2*P(2)+N3*P(3) "*" is a multiplication sign; N1 takes a value of 0 or 1. When the integrated water-falling detection server receives the information A of a person falling into the water, N1 takes a value of 1, otherwise it takes a value of 0; N2 takes a value of 0 or 1. When the comprehensive water-falling detection server receives the information B of a person falling into the water, N2 takes a value of 1, otherwise it takes a value of 0; N3 takes a value of 0 or 1. When the integrated water-falling detection server receives the information C of a person falling into the water, N3 takes a value of 1, otherwise it takes a value of 0; P(1) is the probability of a person actually falling into the water when the comprehensive detection server uses Bayes’ theorem to infer the person falling into the water information A; P(2) is the probability of a person actually falling into the water when the comprehensive detection server uses Bayes’ theorem to infer the person falling into the water information B; P(3) is the probability of a person actually falling into the water when the comprehensive detection server uses Bayes' theorem to infer the probability of a person actually falling into the water when receiving the person falling into the water information C.

10. A 5G-based comprehensive method for detecting people falling into water according to claim 9, characterized in that: Before the comprehensive water detection server uses Bayesian theorem to infer P(1), P(2) and P(3), it includes: Obtain prior probability: obtain the prior probability "P(falling into water)" of a person falling into water in the falling-into-water detection area, the probability "P(receiving person falling-into-water information A)" of the falling-into-water comprehensive detection server receiving person falling-into-water information B "P(receiving person falling-into-water information B)", and the probability "P(receiving person falling-into-water information C)" of the falling-into-water comprehensive detection server receiving person falling-into-water information C; Collect observation data: collect the probability that the comprehensive detection server for falling into water receives the information A of a person falling into the water when a falling-incident occurs, "P(received information about a person falling into the water|falling into the water)", the probability that the comprehensive detection server for falling into water receives the information B of a person falling into the water when a falling-incident occurs, "P(received information about a person falling into the water|falling into the water)", and the probability that the comprehensive detection server for falling into water receives the information C of a person falling into the water when a falling-incident occurs, "P(received information about a person falling into the water|falling into the water)".

Images