A multi-state data labeling method for monitoring ocean dumping
By performing multi-state data marking in the marine dump monitoring system, the problem of insufficient real-time and availability of data in the prior art is solved, effective data classification and abnormal data identification are realized, and law enforcement capabilities are improved.
Patent Information
- Application Number
- CN202411266032.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-11
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-09-11
AI Technical Summary
The existing online marine dumping monitoring technology lacks data marking function, resulting in poor real-time data and low availability, and the inability to effectively distinguish data during abnormal periods and interfere with law enforcement analysis.
A multi-state data marking method is proposed to perform abnormal data processing by obtaining ship data and marking it, including operating state, time, longitude, latitude, speed, draft depth, mud gate state, video and image data. Distinguish between normal and abnormal data, reduce interference and identify illegal behaviors.
The synchronous classification of data and the identification of abnormal data are realized, the data availability and law enforcement and evidence collection capabilities of the monitoring system are improved, and interference with the legality analysis of ship dumping behavior is reduced.
Smart Images

Figure CN118797529B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of marine environmental protection, in particular to a marine dumping monitoring method with multi-state data marking. Background Art
[0002] Ocean dumping refers to the act of disposing of waste into the ocean using carriers such as ships, aircraft, platforms and other artificial structures at sea. Dumped waste includes dredged materials, inert inorganic geological materials, fishery waste, etc. The ocean dumping process is short, illegal markers disappear quickly, and there may be dumping along the way, random dumping, stealing non-dredged materials (construction waste, etc.), and stealing sea sand in the name of dredging. It is necessary to regulate their dumping behavior through ocean dumping online monitoring equipment. There are still certain obstacles to the application of automatic monitoring data generated by current technology to law enforcement. The marking function of data status is almost not considered. The data has poor real-time performance and low availability, which is not conducive to subsequent supervision and law enforcement evidence collection.
[0003] The data collected by the existing technology is not marked. Data marking refers to marking the corresponding time period that generates the automatic monitoring data according to the operation of the automatic monitoring equipment and the data transmission network status to confirm the validity of the automatic monitoring data. Online monitoring of ocean dumping is real-time, online, and uninterrupted. During the objective monitoring process, invalid data generated during the maintenance, failure, and overhaul of the automatic monitoring equipment will be uploaded to the automatic monitoring system. The data generated during these abnormal periods will cause great interference to the analysis and evaluation of whether the operating ships are dumping dredged materials in accordance with the law.
[0004] The invention patent application with application publication number CN115277782A discloses a marine dumping monitoring and management system and method based on GRU and dual CPU, which uses dual CPU, RTOS embedded operating system and Linux embedded operating system to cooperate with each other to realize the collection and processing of different data, simplify the structure, and use GRU neural network model algorithm to comprehensively judge whether there are illegal and irregular behaviors, and automatically record illegal and irregular behaviors and send information to prompt, which can greatly reduce the workload of managers. The disadvantage of this method is that the collected data is not marked, and it cannot distinguish whether it is data generated during abnormal periods. Summary of the invention
[0005] In order to solve the above technical problems, the present invention proposes a method for monitoring ocean dumping with multi-state data marking, which, while collecting conventional data, synchronously collects and marks equipment operation status data, marks data in different situations, and realizes synchronous classification of collected data.
[0006] The object of the present invention is a method for monitoring marine dumping with multi-state data tagging, comprising obtaining vessel data, wherein the vessel data includes operating status, time, longitude, latitude, speed, heading, draft, mud gate status, video and image data, and further comprising the following steps:
[0007] Step 1: extract the mark of the data of the running state, when the mark state is "0", sort the ship data by time as the index, record it as Data, and execute the abnormal data processing method;
[0008] Step 2: Sort the vessel data by time as an index, counted as Data, and associate the engineering and dumping license information corresponding to the vessel with the vessel name and the time of the data as an index;
[0009] Step 3: If the dumping permit information can be linked, make a judgment based on the first mud gate status, the first draft depth and the first track compared with the AIS;
[0010] Step 4: If the dumping permit information cannot be associated, make a judgment based on the second mud gate status, the second draft depth and the comparison of the second track with AIS.
[0011] Preferably, judging the state of the first mud door includes the following sub-steps:
[0012] Step 3001: determine the state of the mud door, and when the mud door is in a closed state, mark it as passed;
[0013] Step 3002: when the mud door is in the open state, filter the data marked as open by the mud door, record it as Open, and define an empty data set, record it as Open_a;
[0014] Step 3003: traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data;
[0015] Step 3004: judging whether Open_ is in the engineering area or the dumping area according to the longitude and latitude, if Open_ is in the engineering area or the dumping area, executing step 3006;
[0016] Step 3005: if Open_ is not in the project area or dumping area, mark Open_ with Alarm_A and count it into Open_a, where Alarm_A is the indicator of dumping not in accordance with the permit;
[0017] Step 3006: When Open_a is empty, it is marked as passed; when Open_a is not empty, it is reported as an alarm information;
[0018] Step 3007: define an empty data set, record it as Open_b; traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data;
[0019] Step 3008: Before calculation n The mean draft of the data is recorded as X 1. After n The mean draft of the data is recorded as Y 1. When X 1- Y 1> a %* X 1, it is judged that the water depth drops suddenly; when X 1- Y 1≤ a %* X 1, mark Open_ as Alarm_B and count it into Open_b, where, a is the water depth judgment threshold, where Alarm_B is the suspected mud gate fault indicator;
[0020] Step 3009: When Open_b is empty, it is marked as passed; when Open_b is not empty, it is reported as an alarm information.
[0021] In any of the above solutions, preferably, judging from the first draft depth includes the following sub-steps:
[0022] Step 3101: define four empty data sets, denoted as Depth, Depth_a, Depth_b and Depth_c; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Data;
[0023] Step 3102: Before calculation n The mean draft of the data is recorded as X 2. After n The mean draft of the data is recorded as Y 2. When X 2- Y 2> a %* X 2, Data_ is included in Depth to determine the sudden drop in water depth; when X 2- Y 2≤ a %* X 2, it is determined that the water depth is constant, and step 3109 is executed;
[0024] Step 3103: traverse Depth, record it as a temporary variable Depth_, and locate the position of Depth_ in Depth;
[0025] Step 3104: judging whether Depth_ is in the engineering area or the dumping area according to the longitude and latitude, if Depth_ is in the engineering area or the dumping area, executing step 3106;
[0026] Step 3105: If Depth_ is not in the engineering area or dumping area, mark Depth_ as Alarm_A and count it into Depth_a;
[0027] Step 3106: When Depth_a is empty, mark it as passed; when Depth_a is not empty, report it as alarm information;
[0028] Step 3107: When the water depth drops suddenly and the mud gate is closed, filter the data with the mud gate marked as closed in Depth, mark Alarm_B and count it in Depth_b;
[0029] Step 3108: When Depth_b is empty, mark it as passed; when Depth_b is not empty, report it as alarm information;
[0030] Step 3109: When X 2- Y When 2=0, the data of ten minutes before and after Data_ is intercepted in Data and recorded as Section; when X 2- Y When 2≠0, execute step 3112;
[0031] Step 3110: Calculate the variance of the position data in Section, record it as list_v, calculate the ratio s of the number of data with value = 0 in list_v to the number of data in list_v,
[0032] Step 3111, judging whether the track has changed according to the ratio s, when s<90%, marking Data_ with Alarm_C and counting it into Depth_c, wherein Alarm_C is a suspected draft sensor failure indicator;
[0033] Step 3112: When Depth_c is empty, it is marked as passed; when Depth_c is not empty, it is reported as an alarm information.
[0034] In any of the above solutions, preferably, judging from the comparison between the first track and the AIS includes the following sub-steps:
[0035] Step 3201: Get the AIS track data of the ship in the same time period as Data, and record it as Dx;
[0036] Step 3202: traverse Data, record it as a temporary variable Data_, and locate Data_ at the position Dx using time as the index;
[0037] Step 3203: define an empty data set, marked as Mark. When the difference between the longitude and latitude data corresponding to Data_ and the longitude and latitude data of Dx at the same time after conversion is greater than 1 km, record Data_ in Mark. When the difference between the longitude and latitude data corresponding to Data_ and the longitude and latitude data of Dx at the same time after conversion is less than or equal to 1 km, mark it as passed.
[0038] Step 3204: define an empty data set, denoted as Mark_a, traverse Mark with time as index, denoted as temporary variable Mark_, when the time interval between Mark_ and the next value is less than m minutes, mark Mark_ as Alarm_E and record it in Mark_a;
[0039] Step 3205: When Mark_a is empty, the mark is passed; when Mark_a is not empty, it is reported as an alarm information.
[0040] In any of the above solutions, preferably, judging the state of the second mud door includes the following sub-steps:
[0041] Step 4001: determine the state of the mud door, and when the mud door is in a closed state, mark it as passed;
[0042] Step 4002: Filter the data marked as open by the mud gate, record it as Open, define an empty data set, record it as Open_c, mark each Open with Alarm_D and count it into Open_c, where Alarm_D is the indicator of unlicensed dumping;
[0043] Step 4003: When Open_c is empty, it is marked as passed; when Open_c is not empty, it is reported as an alarm information;
[0044] Step 4004: define an empty data set, record it as Open_d; traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data;
[0045] Step 4005: Before calculation n The mean draft of the data is recorded as X 3. After n The mean draft of the data is recorded as Y 3. When X 3- Y 3> a %* X 3, it is judged that the water depth drops suddenly; when X 3- Y 3≤ a %* X At 3 o'clock, mark Open_ as Alarm_B and count it into Open_d;
[0046] Step 4006: When Open_b is empty, it is marked as passed; when Open_b is not empty, it is reported as an alarm information.
[0047] In any of the above solutions, preferably, judging from the first draft depth includes the following sub-steps:
[0048] Step 4101: define four empty data sets, denoted as Depth, Depth_x, Depth_y and Depth_z; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Data;
[0049] Step 4102: Before calculation n The mean draft of the data is recorded as X 4. After n The mean draft of the data is recorded as Y 4. When X 4- Y 4> a %* X 4, Data_ is included in Depth to determine the sudden drop in water depth; when X 4- Y 4≤ a %* X 4, it is determined that the water depth is constant, and step 4107 is executed;
[0050] Step 4103: Mark each Depth as Alarm_D and add it to Depth_x;
[0051] Step 4104: When Depth_x is empty, mark it as passed; when Depth_x is not empty, report it as alarm information;
[0052] Step 4105: Filter the data in Depth where the mud gate is marked as closed, mark it as Alarm_B and include it in Depth_y;
[0053] Step 4106: When Depth_y is empty, it is marked as passed; when Depth_y is not empty, it is reported as an alarm information;
[0054] Step 4107: When Y 4- X When 4=0, the data of ten minutes before and after Data_ is intercepted in Data and recorded as Section; when Y 2- X When 2≠0, execute step 4110;
[0055] Step 4108: Calculate the variance of the position data in Section, record it as list_v, and calculate the ratio s of the number of data with value = 0 in list_v to the number of data in list_v;
[0056] Step 4109: judging the track change according to the ratio s, when s<90%, Data_marks Alarm_C and counts it into Depth_z;
[0057] Step 4110: When Depth_z is empty, it is marked as passed; when Depth_z is not empty, it is reported as an alarm information.
[0058] In any of the above solutions, preferably, judging from the comparison between the second track and the AIS includes the following sub-steps:
[0059] Step 4201: Get the ship AIS track data in the same time period as Data, and record it as Dy;
[0060] Step 4202: traverse Data, record it as a temporary variable Data_, and locate Data_ at the position of Dy using time as the index;
[0061] Step 4203: define an empty data set, record it as Mark_1, if the latitude and longitude data corresponding to Data_ differ from the latitude and longitude data of Dy at the same time by more than 1 km after conversion, mark it passed;
[0062] Step 4204: define an empty data set, denoted as Mark_ x, traverse Mark_1 with time as index, denoted as temporary variable Mark_, if the time interval between Mark_ and the next value is less than m minutes, mark Mark_ as Alarm_E and record it in Mark_x;
[0063] Step 4205: When Mark_x is empty, the mark is passed; when Mark_x is not empty, it is reported as an alarm information.
[0064] In any of the above solutions, preferably, the method for obtaining the data with the mark status of "0" is:
[0065] 1) When the dumping monitoring system receives the electrical signal of the backup power supply starting, the data status of the collected data is marked as "O";
[0066] 2) Send an encrypted data collection instruction to the camera, take a photo every 10 seconds, and record its metadata as Pic.
[0067] In any of the above solutions, preferably, the abnormal data processing method includes making judgments based on the third mud gate state and the third draft depth respectively.
[0068] In any of the above solutions, preferably, judging from the third mud door state includes the following sub-steps:
[0069] Step 1001: determine the state of the mud door, and when the mud door is in a closed state, mark it as passed;
[0070] Step 1002: when the mud door is in the open state, filter the data marked as open by the mud door, record it as Open, and define an empty data set, record it as Open_a;
[0071] Step 1003: traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data;
[0072] Step 1004: judging whether Open_ is in the engineering area or the dumping area according to the longitude and latitude, if Open_ is in the engineering area or the dumping area, executing step 1006;
[0073] Step 1005: if Open_ is not in the project area or the dumping area, mark Open_ with Alarm_A and count it into Open_a, where Alarm_A is the indicator of dumping not in accordance with the permit;
[0074] Step 1006: traverse Open_a, record it as a temporary variable Open_a_, record the time of Open_a_ as X1, and the time five minutes after Open_a_ as Y1;
[0075] Step 1007: define an empty data set Pic_M, if the shooting time of Pic is between X1 and Y1, store the data in Pic_M;
[0076] Step 1008: When Open_a is empty, it is marked as passed; when Open_a is not empty, Open_a and Pic_M are reported as alarm information.
[0077] In any of the above solutions, preferably, judging from the third draft depth includes the following sub-steps:
[0078] Step 1101: define four empty data sets, denoted as Depth, Depth_a, Depth_b and Depth_c; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Data;
[0079] Step 1102: Before calculation n The mean draft of the data is recorded as X 5. After n The mean draft of the data is recorded as Y 5;
[0080] Step 1103: When X 5- Y 5> a %* X 5, Data_ is included in Depth to determine the sudden drop in water depth; when X 5- Y 5≤ a %* X 5, it is determined that the water depth is constant, and step 1108 is executed;
[0081] Step 1104: traverse Depth, record it as a temporary variable Depth_, and locate the position of Depth_ in Depth;
[0082] Step 1105: judging whether Depth_ is in the engineering area or the dumping area according to the longitude and latitude, if Depth_ is in the engineering area or the dumping area, marking it as passed; if Depth_ is not in the engineering area or the dumping area, marking Depth_ as Alarm_A and counting it into Depth_a;
[0083] Step 1106: traverse Depth_a, record it as a temporary variable Depth_a_, record the time of Depth_a_ as X2, and the time five minutes after Depth_a_ as Y2;
[0084] Step 1107: define an empty data set Pic_N, if the shooting time of Pic is between X2 and Y2, store the data in Pic_N;
[0085] Step 1108: When Depth_a is empty, mark it as passed; when Depth_a is not empty, report Depth_a and Pic_N as alarm information.
[0086] The present invention proposes a multi-state data marking ocean dumping monitoring method, which classifies the data through data marking while collecting data, wherein normal data can be directly used in the subsequent analysis and judgment of the legality of ship dumping behavior, and abnormal data can be eliminated to reduce interference. In addition, the abnormal data can also be analyzed to identify the illegal behavior of abnormal operation of the ocean dumping online monitoring equipment. BRIEF DESCRIPTION OF THE DRAWINGS
[0087] Figure 1 The present invention is a flowchart of a preferred embodiment of a method for monitoring ocean dumping using multi-state data labeling according to the present invention.
[0088] Figure 2 It is a flow chart of an embodiment of a third mud gate state judgment method of the ocean dumping monitoring method with multi-state data marking according to the present invention.
[0089] Figure 3 It is a flow chart of an embodiment of the third draft depth determination method of the ocean dumping monitoring method with multi-state data marking according to the present invention.
[0090] Figure 4 The present invention is a flowchart of an embodiment of a method for determining the state of a first mud gate according to the method for monitoring ocean dumping with multi-state data marking of the present invention.
[0091] Figure 5 The present invention is a flowchart of an embodiment of a method for comparing and judging the first track and AIS in a method for monitoring ocean dumping with multi-state data marking according to the present invention.
[0092] Figure 6 It is a flow chart of another preferred embodiment of the ocean dumping monitoring method with multi-state data marking according to the present invention. DETAILED DESCRIPTION
[0093] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments.
[0094] Example 1
[0095] like Figure 1 As shown, step 10000 is executed to obtain ship data, where the ship data includes operating status, time, longitude, latitude, speed, heading, draft, mud gate status, video and image data.
[0096] Execute step 11000 to extract the mark of the data of the running state. When the mark state is "0", sort the ship data by time as the index, record it as Data, and execute the abnormal data processing method. The method for obtaining the data with the mark state of "0" is:
[0097] 1) When the dumping monitoring system receives the electrical signal of the backup power supply starting, the data status of the collected data is marked as "O";
[0098] 2) Send an encrypted data collection instruction to the camera, take a photo every 10 seconds, and record its metadata as Pic.
[0099] The abnormal data processing method includes making judgments based on the third mud gate state and the third draft depth respectively.
[0100] 1. If Figure 2 As shown, judging the state of the third mud door includes the following sub-steps:
[0101] Step 11001: determine the state of the mud door, and when the mud door is in a closed state, mark it as passed;
[0102] Step 11002: when the mud door is in the open state, filter the data marked as open by the mud door, record it as Open, and define an empty data set, record it as Open_a;
[0103] Step 11003: traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data;
[0104] Step 11004: judging whether Open_ is in the engineering area or the dumping area according to the longitude and latitude, if Open_ is in the engineering area or the dumping area, executing step 11006;
[0105] Step 11005: if Open_ is not in the project area or dumping area, mark Open_ with Alarm_A and count it into Open_a, where Alarm_A is the indicator of dumping not in accordance with the permit;
[0106] Step 11006: traverse Open_a, record it as a temporary variable Open_a_, record the time of Open_a_ as X1, and the time five minutes after Open_a_ as Y1;
[0107] Step 11007: define an empty data set Pic_M. If the shooting time of Pic is between X1 and Y1, store the data in Pic_M.
[0108] Step 11008: When Open_a is empty, it is marked as passed; when Open_a is not empty, Open_a and Pic_M are reported as alarm information.
[0109] 2. If Figure 3 As shown, judging from the third draft depth includes the following sub-steps:
[0110] Step 11101: define four empty data sets, denoted as Depth, Depth_a, Depth_b and Depth_c; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Data;
[0111] Step 11102: Before calculation n The mean draft of the data is recorded as X 5. After n The mean draft of the data is recorded as Y 5;
[0112] Step 11103: When X 5- Y 5> a %* X 5, Data_ is included in Depth to determine the sudden drop in water depth; whenX 5- Y 5≤ a %* X 5, it is determined that the water depth is constant, and step 11108 is executed;
[0113] Step 11104: traverse Depth, record it as a temporary variable Depth_, and locate the position of Depth_ in Depth;
[0114] Step 11105: Determine whether Depth_ is in the engineering area or the dumping area according to the longitude and latitude. If Depth_ is in the engineering area or the dumping area, mark it as passed; if Depth_ is not in the engineering area or the dumping area, mark Depth_ as Alarm_A and count it into Depth_a;
[0115] Step 11106: traverse Depth_a, record it as a temporary variable Depth_a_, record the time of Depth_a_ as X2, and the time five minutes after Depth_a_ as Y2;
[0116] Step 11107: define an empty data set Pic_N, if the shooting time of Pic is between X2 and Y2, store the data in Pic_N;
[0117] Step 11108: When Depth_a is empty, mark it as passed; when Depth_a is not empty, report Depth_a and Pic_N as alarm information.
[0118] Execute step 12000, sort the vessel data by time as index, count as Data, use the vessel name and the time of the data as index, and associate the engineering and dumping license information corresponding to the vessel.
[0119] Execute step 13000. If the dumping permit information can be associated, make a judgment based on the first mud gate status, the first draft depth, and the comparison between the first track and AIS.
[0120] 1. If Figure 4 As shown, judging the state of the first mud door includes the following sub-steps:
[0121] Step 13001: determine the state of the mud door, and when the mud door is in a closed state, mark it as passed;
[0122] Step 13002: when the mud door is in the open state, filter the data marked as open by the mud door, record it as Open, and define an empty data set, record it as Open_a;
[0123] Step 13003: traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data;
[0124] Step 13004: judging whether Open_ is in the engineering area or the dumping area according to the longitude and latitude, if Open_ is in the engineering area or the dumping area, executing step 13006;
[0125] Step 13005: If Open_ is not in the project area or dumping area, mark Open_ with Alarm_A and count it into Open_a, where Alarm_A is the indicator of dumping not in accordance with the permit;
[0126] Step 13006: When Open_a is empty, mark it as passed; when Open_a is not empty, report it as alarm information;
[0127] Step 13007: define an empty data set, record it as Open_b; traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data;
[0128] Step 13008: Before calculation n The mean draft of the data is recorded as X 1. After n The mean draft of the data is recorded as Y 1. When X 1- Y 1> a %* X 1, it is judged that the water depth drops suddenly; when X 1- Y 1≤ a %* X 1, mark Open_ as Alarm_B and count it into Open_b, where, a is the water depth judgment threshold, where Alarm_B is the suspected mud gate fault indicator;
[0129] Step 13009: When Open_b is empty, it is marked as passed; when Open_b is not empty, it is reported as an alarm information.
[0130] 2. Judging from the first draft depth includes the following sub-steps:
[0131] Step 13101: define four empty data sets, denoted as Depth, Depth_a, Depth_b and Depth_c; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Data;
[0132] Step 13102: Before calculation nThe mean draft of the data is recorded as X 2. After n The mean draft of the data is recorded as Y 2;
[0133] Step 13103: When X 2- Y 2> a %* X 2, Data_ is included in Depth to determine the sudden drop in water depth; when X 2- Y 2≤ a %* X 2, it is determined that the water depth is constant, and step 13110 is executed;
[0134] Step 13104: traverse Depth, record it as a temporary variable Depth_, and locate the position of Depth_ in Depth;
[0135] Step 13105: judging whether Depth_ is in the engineering area or the dumping area according to the longitude and latitude, if Depth_ is in the engineering area or the dumping area, executing step 13107;
[0136] Step 13106: If Depth_ is not in the engineering area or dumping area, mark Depth_ as Alarm_A and include it in Depth_a;
[0137] Step 13107: When Depth_a is empty, mark it as passed; when Depth_a is not empty, report it as alarm information;
[0138] Step 13108: When the water depth drops suddenly and the mud gate is closed, filter the data with the mud gate marked as closed in Depth, mark Alarm_B and count it in Depth_b;
[0139] Step 13109: When Depth_b is empty, mark it as passed; when Depth_b is not empty, report it as alarm information;
[0140] Step 13110: When X 2- Y When 2=0, the data of ten minutes before and after Data_ is intercepted in Data and recorded as Section; when X 2- Y When 2≠0, execute step 13113;
[0141] Step 13111: Calculate the variance of the position data in Section, record it as list_v, and calculate the ratio s of the number of data with value = 0 in list_v to the number of data in list_v;
[0142] Step 13112, judging whether the track has changed according to the ratio s, when s<90%, marking Data_ with Alarm_C and counting it into Depth_c, wherein Alarm_C is a suspected draft sensor failure indicator;
[0143] Step 13113: When Depth_c is empty, it is marked as passed; when Depth_c is not empty, it is reported as an alarm information.
[0144] 3. If Figure 5 As shown, judging by comparing the first track with the AIS includes the following sub-steps:
[0145] Step 13201: Get the ship AIS track data in the same time period as Data, and record it as Dx;
[0146] Step 13202: traverse Data, record it as a temporary variable Data_, and locate Data_ at the position Dx using time as the index;
[0147] Step 13203: define an empty data set, denoted as Mark;
[0148] Step 13204: When the difference between the longitude and latitude data corresponding to Data_ and the longitude and latitude data of Dx at the same time after conversion is greater than 1 kilometer, record Data_ in Mark; when the difference between the longitude and latitude data corresponding to Data_ and the longitude and latitude data of Dx at the same time after conversion is less than or equal to 1 kilometer, execute step 13206;
[0149] Step 13205: define an empty data set, denoted as Mark_a, traverse Mark with time as index, denoted as temporary variable Mark_, when the time interval between Mark_ and the next value is less than m minutes, mark Mark_ as Alarm_E and record it in Mark_a;
[0150] Step 13206: When Mark_a is empty, the mark is passed; when Mark_a is not empty, it is reported as an alarm message.
[0151] Execute step 14000. If the dumping permit information cannot be associated, make a judgment based on the second mud gate status, the second draft depth, and the comparison between the second track and AIS.
[0152] 1. Judging the state of the second mud door includes the following sub-steps:
[0153] Step 14001: determine the state of the mud door, and when the mud door is in a closed state, mark it as passed;
[0154] Step 14002: Filter the data marked as open for the mud gate, record it as Open, define an empty data set, record it as Open_c, mark each Open with Alarm_D and count it into Open_c, where Alarm_D is the indicator of unlicensed dumping;
[0155] Step 14003: When Open_c is empty, it is marked as passed; when Open_c is not empty, it is reported as an alarm information;
[0156] Step 14004: define an empty data set, record it as Open_d; traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data;
[0157] Step 14005: Before calculation n The mean draft of the data is recorded as X 3. After n The mean draft of the data is recorded as Y 3. When X 3- Y 3> a %* X 3, it is judged that the water depth drops suddenly; when X 3- Y 3≤ a %* X At 3 o'clock, mark Open_ as Alarm_B and count it into Open_d;
[0158] Step 14006: When Open_b is empty, it is marked as passed; when Open_b is not empty, it is reported as an alarm information.
[0159] 2. Judging from the second draft depth includes the following sub-steps:
[0160] Step 14101: define four empty data sets, denoted as Depth, Depth_x, Depth_y and Depth_z; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Data;
[0161] Step 14102: Before calculation n The mean draft of the data is recorded as X 4. After n The mean draft of the data is recorded as Y 4. When X 4- Y 4> a %* X 4, Data_ is included in Depth to determine the sudden drop in water depth; when X 4- Y 4≤a %* X 4, it is determined that the water depth is constant, and step 14107 is executed;
[0162] Step 14103: Mark each Depth as Alarm_D and add it to Depth_x;
[0163] Step 14104: When Depth_x is empty, mark it as passed; when Depth_x is not empty, report it as alarm information;
[0164] Step 14105: Filter the data in Depth where the mud gate is marked as closed, mark Alarm_B and include it in Depth_y;
[0165] Step 14106: When Depth_y is empty, it is marked as passed; when Depth_y is not empty, it is reported as an alarm information;
[0166] Step 14107: When Y 4- X When 4=0, the data of ten minutes before and after Data_ is intercepted in Data and recorded as Section; when Y 2- X When 2≠0, execute step 14110;
[0167] Step 14108: Calculate the variance of the position data in Section, record it as list_v, and calculate the ratio s of the number of data with value = 0 in list_v to the number of data in list_v;
[0168] Step 14109: judging the track change according to the ratio s, when s<90%, Data_marks Alarm_C and counts it into Depth_z;
[0169] Step 14110: When Depth_z is empty, it is marked as passed; when Depth_z is not empty, it is reported as an alarm information.
[0170] 3. Comparing the second track with the AIS includes the following sub-steps:
[0171] Step 14201: Get the ship AIS track data in the same time period as Data, and record it as Dy;
[0172] Step 14202: traverse Data, record it as a temporary variable Data_, and locate Data_ at the position of Dy using time as the index;
[0173] Step 14203: define an empty data set, record it as Mark_1, if the latitude and longitude data corresponding to Data_ differ from the latitude and longitude data of Dy at the same time by more than 1 km after conversion, mark it passed;
[0174] Step 14204: define an empty data set, denoted as Mark_ x, traverse Mark_1 with time as index, denoted as temporary variable Mark_, if the time interval between Mark_ and the next value is less than m minutes, mark Mark_ as Alarm_E and record it in Mark_x;
[0175] Step 14205: When Mark_x is empty, the mark is passed; when Mark_x is not empty, it is reported as an alarm message.
[0176] Example 2
[0177] In order to overcome the problem that existing online monitoring equipment for ocean dumping cannot effectively classify data and cannot present data generated during abnormal periods, the present invention provides an ocean dumping monitoring method and system that provides multi-state data marking. While having the function of collecting conventional data, it can also synchronously collect and mark equipment operating status data, mark data in different situations, and achieve synchronous classification of collected data.
[0178] like Figure 6 As shown, the following steps are included:
[0179] Step 1: Take out a piece of data containing time, location, draft, and mud gate, sort it by time, record it as Date, use the ship name and data time as index, associate the project and dumping permit information corresponding to the ship, if the dumping permit information can be associated, execute step 2, if the dumping permit information cannot be associated, execute step 27.
[0180] Step 2 (mud gate open): Filter the data with the mud gate marked as open and record it as Open; define an empty data set and record it as Open_a.
[0181] Step 3: Traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Date.
[0182] Step 4 (whether it is located in the engineering area or dumping area): Determine whether Open_ is in the engineering area qdgc based on the latitude and longitude data. If yes, pass; if not, proceed to the next step.
[0183] Step 5: Determine whether Open_ is in the dumping area based on the latitude and longitude data. If yes, pass; if not, mark Open_ as Alarm_A and count it into Open_a.
[0184] Step 6: If Open_a is not empty, it is reported as an alarm message.
[0185] Step 7 (determine the draft depth after the mud gate is opened): define an empty data set, recorded as Open_b; traverse Open, recorded as a temporary variable Open_, and locate the position of Open_ in Date.
[0186] Step 8 (mud gate is open and draft drops suddenly): calculate the average draft of the first n data and record it as X, and the average draft of the last n data and record it as Y; if XY>a%*X, then pass, otherwise mark Open_ with Alarm_B and count it in Open_b; a and n are parameters for judging whether there is a sudden drop in water depth, which shall be determined according to the actual situation, the same below.
[0187] Step 9: If Open_b is not empty, it is reported as an alarm message.
[0188] Step 10 (Determine the draft depth): Define four empty data sets, denoted as Depth, Depth_a, Depth_b, and Depth_c; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Date.
[0189] Step 11: Calculate the mean draft of the first n data, record it as X, and the mean draft of the last n data, record it as Y.
[0190] Step 12 (draft drop): If XY>a%*X, then include Data_ in Depth, otherwise proceed to step 20.
[0191] Step 13: Traverse Depth, record it as a temporary variable Depth_, and locate the position of Depth_ in Depth.
[0192] Step 14: Determine whether Depth_ is within the project area qdgc based on the latitude and longitude data. If yes, pass; if not, proceed to the next step. a and n are parameters for determining whether there is a sudden drop in water depth, which should be determined based on actual conditions.
[0193] Step 15: Determine whether Depth_ is within the dumping area based on the latitude and longitude data. If so, pass; if not, mark Depth_ as Alarm_A and include it in Depth_a.
[0194] Step 16: If Depth_a is not empty, it is reported as an alarm information.
[0195] Step 17 (draft drops suddenly and mud gates are closed): Filter the data in Depth where mud gates are marked as closed, mark Alarm_B and include it in Depth_b.
[0196] Step 18: If Depth_b is not empty, it is reported as an alarm information.
[0197] Step 19 (constant draft): If XY=0, extract the data of ten minutes before and after Data_ in Data and record it as Section; otherwise pass.
[0198] Step 20 (constant draft track remains unchanged): Calculate the variance of the position data in Section, record it as list_v, calculate the ratio of the number of data with value = 0 in list_v to the number of data in list_v, record it as s, s is the parameter to judge whether the track of this group of data remains unchanged, if the track remains unchanged, pass, otherwise mark Data_ as Alarm_C and count it into Depth_c.
[0199] Step 21: If Depth_c is not empty, it is reported as an alarm information.
[0200] Step 22 (track and AIS comparison): Get the ship AIS track data in the same time period as Data, and record it as Dx.
[0201] Step 23: Traverse Date and record it as a temporary variable Date_, and use time as the index to locate the position of Date_ at Dx.
[0202] Step 24: Define an empty data set, named Mark. If the latitude and longitude data corresponding to Date_ differ from the latitude and longitude data of Dx at the same time by more than 1 km after conversion, record Date_ in Mark, otherwise pass.
[0203] Step 25: Define an empty data set, denoted as Mark_a, and traverse Mark with time as the index, denoted as the temporary variable Mark_. If the time interval between Mark_ and the next value is less than m minutes, mark Mark_ with Alarm_E and record it in Mark_a, otherwise pass, where m is a parameter to determine whether the track is continuous.
[0204] Step 26: If Mark_a is not empty, generate alarm data for reporting.
[0205] The following are the ones that failed to associate the license information.
[0206] Step 27 (mud gate open): Filter the data marked as open for the mud gate, record it as Open, define an empty data set, record it as Open_c, mark each Open with Alarm_D and count it into Open_c.
[0207] Step 28: If Open_c is not empty, it is reported as an alarm message.
[0208] Step 29 (determine the draft depth after the mud gate is opened): define an empty data set, recorded as Open_d; traverse Open, recorded as a temporary variable Open_, and locate the position of Open_ in Date.
[0209] Step 30 (mud gate is open and draft drops suddenly): calculate the average draft of the first n data and record it as X, and the average draft of the last n data and record it as Y; if XY>a%*X, then pass, otherwise mark Open_ with Alarm_B and count it into Open_d; where a and n are parameters for judging whether there is a sudden drop in water depth, which shall be determined according to the actual situation, the same below.
[0210] Step 31: If Open_d is not empty, it is reported as an alarm information.
[0211] Step 32 (determine draft depth): define four empty data sets, denoted as Depth, Depth_x, Depth_y and Depth_z; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Date.
[0212] Step 33: Calculate the mean draft of the first n data, record it as X, and the mean draft of the last n data, record it as Y.
[0213] Step 34 (draft drop): If XY>a%*X, then include Data_ in Depth, otherwise go to step 39.
[0214] Step 35: Mark each Depth as Alarm_D and include it in Depth_x.
[0215] Step 36: If Depth_x is not empty, it is reported as an alarm message.
[0216] Step 37 (draft drops suddenly and mud gate is closed): Filter the data in Depth where the mud gate is marked as closed, mark Alarm_B and include it in Depth_y.
[0217] Step 38: If Depth_y is not empty, it is reported as an alarm information.
[0218] Step 39 (constant draft): If XY=0, extract the data of ten minutes before and after Data_ in Data and record it as Section; otherwise pass.
[0219] Step 40 (constant draft track remains unchanged): Calculate the variance of the position data in Section, record it as list_v, calculate the ratio of the number of data with value = 0 in list_v to the number of data in list_v, record it as s, s is the parameter for judging whether the track of this group of data remains unchanged, if the track remains unchanged, pass, otherwise mark Data_ with Alarm_C and count it into Depth_z.
[0220] Step 41: If Depth_z is not empty, it is reported as an alarm information.
[0221] Step 42 (track and AIS comparison): Get the ship AIS track data in the same time period as Data, and record it as Dy.
[0222] Step 43: Traverse Date and record it as a temporary variable Date_, and use time as an index to locate the position of Date_ at Dy.
[0223] Step 44: Define an empty data set, record it as Mark_1. If the latitude and longitude data corresponding to Date_ differ from the latitude and longitude data of the same time Dy by more than 1 km after conversion, record Date_ in Mark_1, otherwise pass.
[0224] Step 45: Define an empty data set, denoted as Mark_x, and traverse Mark_1 with time as the index, denoted as the temporary variable Mark_. If the time interval between Mark_ and the next value is less than m minutes, mark Mark_ with Alarm_E and record it in Mark_x, otherwise pass. m is a parameter to determine whether the track is continuous.
[0225] Step 46: If Mark_x is not empty, generate alarm data for reporting.
[0226] The alarm information has been corresponding to different data tags, as shown in Table 1.
[0227] Table 1 Data label corresponding situation table
[0228]
[0229] The present invention collects data and classifies the data through data marking, wherein normal data can be directly used in subsequent analysis and judgment on the legality of ship dumping behavior, and abnormal data can be eliminated to reduce interference. In addition, abnormal data can also be analyzed to identify illegal behavior of abnormal operation of ocean dumping online monitoring equipment.
[0230] Example 3
[0231] On the basis of the original dumping system, the marine dumping monitoring system proposed by the present invention adds a backup power supply, a Hall sensor and multiple backup dumping collection cameras. The use of the marine dumping monitoring system is monitored by the Hall sensor. When the Hall sensor captures that the backup power supply is enabled, the collected data is automatically marked as the code "O" corresponding to "power off" and the backup dumping collection camera is started to collect video data until the power supply of the equipment returns to normal. The collected "power off" data is used in an abnormal data monitoring method, which includes the following steps:
[0232] Step 1: When receiving the electrical signal indicating that the backup power supply is started, the data state of the collected data is marked as "O".
[0233] Step 2: Send an encrypted data collection instruction to the camera, take a picture every 10 seconds, and record its metadata as Pic.
[0234] Step 3: For data with a data mark status of "O", sort them by time and record them as Data.
[0235] Step 4 (mud gate open): Filter the data with the mud gate marked as open and record it as Open; define an empty data set and record it as Open_a.
[0236] Step 5: Traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data.
[0237] Step 6 (whether it is located in the engineering area or dumping area): Determine whether Open_ is in the engineering area qdgc based on the latitude and longitude data. If yes, pass; if not, proceed to the next step.
[0238] Step 7: Determine whether Open_ is in the dumping area based on the latitude and longitude data. If yes, pass; if not, mark Open_ as Alarm_A and count it into Open_a.
[0239] Step 8: Traverse Open_a and record it as a temporary variable Open_ a_. Record the time of Open_ a_ as X1 and the time five minutes after Open_ a_ as Y1.
[0240] Step 9: Define an empty data set Pic_M. If the shooting time of Pic is between X1 and Y1, store the data in Pic_M.
[0241] Step 10: If Open_a is not empty, Open_a and Pic_M are reported as alarm information.
[0242] Step 11 (determine the draft depth): define four empty data sets, denoted as Depth, Depth_a, Depth_b and Depth_c; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Data.
[0243] Step 12: Calculate the mean draft of the first n data, record it as X, and the mean draft of the last n data, record it as Y.
[0244] Step 13 (draft drop): If XY>a%*X, then include Data_ in Depth, otherwise proceed to step 20.
[0245] Step 14: Traverse Depth, record it as a temporary variable Depth_, and locate the position of Depth_ in Depth.
[0246] Step 15: Determine whether Depth_ is within the project area qdgc based on the latitude and longitude data. If yes, pass; if not, proceed to the next step; a and n are parameters for determining whether there is a sudden drop in water depth, which are determined according to actual conditions.
[0247] Step 16: Determine whether Depth_ is within the dumping area based on the latitude and longitude data. If yes, pass; if not, mark Depth_ as Alarm_A and include it in Depth_a.
[0248] Step 17: Traverse Depth_a and record it as a temporary variable Depth_a_. Record the time of Depth_a_ as X2, and the time five minutes after Depth_a_ as Y2.
[0249] Step 18: Define an empty data set Pic_N. If the shooting time of Pic is between X2 and Y2, store the data in Pic_N.
[0250] Step 19: If Depth_a is not empty, Depth_a and Pic_N are reported as alarm information.
[0251] In order to better understand the present invention, the above is described in detail in conjunction with the specific embodiments of the present invention, but it is not intended to limit the present invention. Any simple modifications made to the above embodiments based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention. Each embodiment in this specification focuses on the differences from other embodiments, and the same or similar parts between the embodiments can be referenced to each other. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
Claims
1. A method for monitoring marine dumping with multi-state data labeling, comprising acquiring ship data, wherein the ship data includes operating status, time, longitude, latitude, speed, heading, draft, mud gate status, video and image data, and synchronously collecting and labeling equipment operating status data while collecting conventional data, labeling data of different situations, and realizing synchronous classification of collected data, characterized in that: The following steps are also included: Step 1: Extract the mark of the data of the running status. When the mark state is "0", sort the ship data by time as the index, record it as Data, and execute the abnormal data processing method. The method for obtaining the data with the mark state of "0" is: 1) When the dumping monitoring system receives the electrical signal of the backup power supply starting, the data status of the collected data is marked as "O"; 2) Send an instruction to the camera to encrypt and collect data, take a picture every 10 seconds, and record its metadata as Pic; Step 2: Sort the vessel data by time as an index, counted as Data, and associate the engineering and dumping license information corresponding to the vessel with the vessel name and the time of the data as an index; Step 3: If the dumping permit information can be linked, make a judgment based on the first mud gate status, the first draft depth and the first track compared with the AIS; Step 4: If the dumping permit information cannot be associated, make a judgment based on the second mud gate status, the second draft depth and the comparison of the second track with AIS.
2. The method for monitoring ocean dumping with multi-state data marking according to claim 1, characterized in that: Judging from the first mud door state includes the following sub-steps: Step 3001: determine the state of the mud door, and when the mud door is in a closed state, mark it as passed; Step 3002: when the mud door is in the open state, filter the data marked as open by the mud door, record it as Open, and define an empty data set, record it as Open_a; Step 3003: traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data; Step 3004: judging whether Open_ is in the engineering area or the dumping area according to the longitude and latitude, if Open_ is in the engineering area or the dumping area, executing step 3006; Step 3005: if Open_ is not in the project area or dumping area, mark Open_ with Alarm_A and count it into Open_a, where Alarm_A is the indicator of dumping not in accordance with the permit; Step 3006: When Open_a is empty, mark it as passed; when Open_a is not empty, report it as alarm information; Step 3007: define an empty data set, record it as Open_b; traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data; Step 3008: Before calculation n The mean draft of the data is recorded as X 1. After n The mean draft of the data is recorded as Y 1. When X 1- Y 1> a %* X 1, it is judged that the water depth drops suddenly; when X 1- Y 1≤ a %* X 1, mark Open_ as Alarm_B and count it into Open_b, where, a is the water depth judgment threshold, where Alarm_B is the suspected mud gate fault indicator; Step 3009: When Open_b is empty, it is marked as passed; when Open_b is not empty, it is reported as an alarm information.
3. The method for monitoring ocean dumping with multi-state data marking as claimed in claim 2, characterized in that: Judging from the first draft depth includes the following sub-steps: Step 3101: define four empty data sets, denoted as Depth, Depth_a, Depth_b and Depth_c; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Data; Step 3102: Before calculation n The mean draft of the data is recorded as X 2. After n The mean draft of the data is recorded as Y 2. When X 2- Y 2> a %* X 2, Data_ is included in Depth to determine the sudden drop in water depth; when X 2- Y 2≤ a %* X 2, it is determined that the water depth is constant, and step 3109 is executed; Step 3103: traverse Depth, record it as a temporary variable Depth_, and locate the position of Depth_ in Depth; Step 3104: judging whether Depth_ is in the engineering area or the dumping area according to the longitude and latitude, if Depth_ is in the engineering area or the dumping area, executing step 3106; Step 3105: If Depth_ is not in the engineering area or dumping area, mark Depth_ as Alarm_A and count it into Depth_a; Step 3106: When Depth_a is empty, mark it as passed; when Depth_a is not empty, report it as alarm information; Step 3107: When the water depth drops suddenly and the mud gate is closed, filter the data with the mud gate marked as closed in Depth, mark Alarm_B and count it in Depth_b; Step 3108: When Depth_b is empty, mark it as passed; when Depth_b is not empty, report it as alarm information; Step 3109: When X 2- Y When 2=0, the data of ten minutes before and after Data_ is intercepted in Data and recorded as Section; when X 2- Y When 2≠0, execute step 3112; Step 3110: Calculate the variance of the position data in Section, record it as list_v, and calculate the ratio s of the number of data with value = 0 in list_v to the number of data in list_v; Step 3111, judging whether the track has changed according to the ratio s, when s<90%, marking Data_ with Alarm_C and counting it into Depth_c, wherein Alarm_C is a suspected draft sensor failure indicator; Step 3112: When Depth_c is empty, it is marked as passed; when Depth_c is not empty, it is reported as an alarm information.
4. The method for monitoring ocean dumping with multi-state data marking according to claim 3, characterized in that: The comparison and judgment between the first track and the AIS includes the following sub-steps: Step 3201: Get the AIS track data of the ship in the same time period as Data, and record it as Dx; Step 3202: traverse Data, record it as a temporary variable Data_, and locate Data_ at the position Dx using time as the index; Step 3203: define an empty data set, marked as Mark. When the difference between the longitude and latitude data corresponding to Data_ and the longitude and latitude data of Dx at the same time after conversion is greater than 1 km, record Data_ in Mark. When the difference between the longitude and latitude data corresponding to Data_ and the longitude and latitude data of Dx at the same time after conversion is less than or equal to 1 km, mark it as passed. Step 3204: define an empty data set, denoted as Mark_a, traverse Mark with time as index, denoted as temporary variable Mark_, when the time interval between Mark_ and the next value is less than m minutes, mark Mark_ as Alarm_E and record it in Mark_a; Step 3205: When Mark_a is empty, the mark is passed; when Mark_a is not empty, it is reported as an alarm information.
5. The method for monitoring ocean dumping with multi-state data marking according to claim 4, characterized in that: Judging from the second mud door state includes the following sub-steps: Step 4001: determine the state of the mud door, and when the mud door is in a closed state, mark it as passed; Step 4002: Filter the data marked as open by the mud gate, record it as Open, define an empty data set, record it as Open_c, mark each Open with Alarm_D and count it into Open_c, where Alarm_D is the indicator of unlicensed dumping; Step 4003: When Open_c is empty, it is marked as passed; when Open_c is not empty, it is reported as an alarm information; Step 4004: define an empty data set, record it as Open_d; traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data; Step 4005: Before calculation n The mean draft of the data is recorded as X 3. After n The mean draft of the data is recorded as Y 3. When X 3- Y 3> a %* X 3, it is judged that the water depth drops suddenly; when X 3- Y 3≤ a %* X At 3 o'clock, mark Open_ as Alarm_B and count it into Open_d; Step 4006: When Open_b is empty, it is marked as passed; when Open_b is not empty, it is reported as an alarm information.
6. The method for monitoring ocean dumping with multi-state data marking according to claim 5, characterized in that: Judging from the second draft depth includes the following sub-steps: Step 4101: define four empty data sets, denoted as Depth, Depth_x, Depth_y and Depth_z; traverse Data, denoted as temporary variable Data_, and locate the position of Data_ in Data; Step 4102: Before calculation n The mean draft of the data is recorded as X 4. After n The mean draft of the data is recorded as Y 4. When X 4- Y 4> a %* X 4, Data_ is included in Depth to determine the sudden drop in water depth; when X 4- Y 4≤ a %* X 4, it is determined that the water depth is constant, and step 4107 is executed; Step 4103: Mark each Depth as Alarm_D and add it to Depth_x; Step 4104: When Depth_x is empty, mark it as passed; when Depth_x is not empty, report it as alarm information; Step 4105: Filter the data in Depth where the mud gate is marked as closed, mark it as Alarm_B and include it in Depth_y; Step 4106: When Depth_y is empty, it is marked as passed; when Depth_y is not empty, it is reported as an alarm information; Step 4107: When Y 4- X When 4=0, the data of ten minutes before and after Data_ is intercepted in Data and recorded as Section; when Y 2- X When 2≠0, execute step 4112; Step 4108: Calculate the variance of the position data in Section, record it as list_v, and calculate the ratio s of the number of data with value = 0 in list_v to the number of data in list_v; Step 4109: judging the track change according to the ratio s, when s<90%, Data_marks Alarm_C and counts it into Depth_z; Step 4110: When Depth_z is empty, it is marked as passed; when Depth_z is not empty, it is reported as an alarm information.
7. The method for monitoring ocean dumping with multi-state data marking according to claim 6, characterized in that: The comparison and judgment between the second track and the AIS includes the following sub-steps: Step 4201: Get the ship AIS track data in the same time period as Data, and record it as Dy; Step 4202: traverse Data, record it as a temporary variable Data_, and locate the position of Data_ at Dy using time as an index; Step 4203: define an empty data set, record it as Mark_1, if the latitude and longitude data corresponding to Data_ differ from the latitude and longitude data of Dy at the same time by more than 1 km after conversion, mark it passed; Step 4204: define an empty data set, denoted as Mark_ x, traverse Mark_1 with time as index, denoted as temporary variable Mark_, if the time interval between Mark_ and the next value is less than m minutes, mark Mark_ as Alarm_E and record it in Mark_x; Step 4206: When Mark_x is empty, the mark is passed; when Mark_x is not empty, it is reported as an alarm information.
8. The method for monitoring ocean dumping with multi-state data marking according to claim 7, characterized in that: The abnormal data processing method includes making judgments based on the third mud gate state and the third draft depth respectively.
9. The method for monitoring ocean dumping with multi-state data marking according to claim 8, characterized in that: Judging from the state of the third mud door includes the following sub-steps: Step 1001: determine the state of the mud door, and when the mud door is in a closed state, mark it as passed; Step 1002: when the mud door is in the open state, filter the data marked as open by the mud door, record it as Open, and define an empty data set, record it as Open_a; Step 1003: traverse Open, record it as a temporary variable Open_, and locate the position of Open_ in Data; Step 1004: judging whether Open_ is in the engineering area or the dumping area according to the longitude and latitude, if Open_ is in the engineering area or the dumping area, executing step 1006; Step 1005: if Open_ is not in the engineering area or the dumping area, mark Open_ with Alarm_A and count it into Open_a, where Alarm_A is the indicator of dumping not in accordance with the permit; Step 1006: Traverse Open_a, record it as a temporary variable Open_a_, record the time of Open_a_ as X1, and the time five minutes after Open_a_ as Y1; Step 1007: define an empty data set Pic_M, if the shooting time of Pic is between X1 and Y1, store the data in Pic_M; Step 1008: When Open_a is empty, it is marked as passed; when Open_a is not empty, Open_a and Pic_M are reported as alarm information.
