WO2024104461A1 - 一种大型船舶安全监管系统 - Google Patents
一种大型船舶安全监管系统 Download PDFInfo
- Publication number
- WO2024104461A1 WO2024104461A1 PCT/CN2023/132323 CN2023132323W WO2024104461A1 WO 2024104461 A1 WO2024104461 A1 WO 2024104461A1 CN 2023132323 W CN2023132323 W CN 2023132323W WO 2024104461 A1 WO2024104461 A1 WO 2024104461A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- ship
- crew
- monitoring
- layer
- Prior art date
Links
- 238000012544 monitoring process Methods 0.000 claims abstract description 58
- 230000005540 biological transmission Effects 0.000 claims abstract description 55
- 238000007906 compression Methods 0.000 claims abstract description 47
- 230000006835 compression Effects 0.000 claims abstract description 47
- 238000007726 management method Methods 0.000 claims abstract description 45
- 238000005516 engineering process Methods 0.000 claims abstract description 43
- 238000013528 artificial neural network Methods 0.000 claims abstract description 40
- 238000000034 method Methods 0.000 claims abstract description 39
- 230000036541 health Effects 0.000 claims abstract description 36
- 238000004458 analytical method Methods 0.000 claims abstract description 32
- 238000010276 construction Methods 0.000 claims abstract description 31
- 238000005457 optimization Methods 0.000 claims abstract description 23
- 230000006870 function Effects 0.000 claims abstract description 21
- 230000003044 adaptive effect Effects 0.000 claims abstract description 12
- 230000036760 body temperature Effects 0.000 claims abstract description 8
- 230000036772 blood pressure Effects 0.000 claims abstract description 7
- 238000004422 calculation algorithm Methods 0.000 claims description 24
- 238000004891 communication Methods 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 18
- 210000002569 neuron Anatomy 0.000 claims description 15
- 230000002159 abnormal effect Effects 0.000 claims description 9
- 238000009529 body temperature measurement Methods 0.000 claims description 9
- 238000004519 manufacturing process Methods 0.000 claims description 9
- 238000005259 measurement Methods 0.000 claims description 9
- 238000010295 mobile communication Methods 0.000 claims description 9
- 238000012550 audit Methods 0.000 claims description 8
- 238000013145 classification model Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000012669 compression test Methods 0.000 claims description 6
- 230000004927 fusion Effects 0.000 claims description 6
- 238000009987 spinning Methods 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000013461 design Methods 0.000 claims description 5
- 230000007613 environmental effect Effects 0.000 claims description 5
- 238000007619 statistical method Methods 0.000 claims description 5
- 238000001931 thermography Methods 0.000 claims description 5
- 229910000831 Steel Inorganic materials 0.000 claims description 4
- 238000007689 inspection Methods 0.000 claims description 4
- 239000010959 steel Substances 0.000 claims description 4
- 238000003860 storage Methods 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 230000000007 visual effect Effects 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 3
- 239000008280 blood Substances 0.000 claims description 3
- 210000004369 blood Anatomy 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 229910052760 oxygen Inorganic materials 0.000 claims description 3
- 239000001301 oxygen Substances 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 claims description 2
- 238000003745 diagnosis Methods 0.000 claims description 2
- 239000011159 matrix material Substances 0.000 claims description 2
- 231100000817 safety factor Toxicity 0.000 claims description 2
- 238000012795 verification Methods 0.000 claims description 2
- 230000003862 health status Effects 0.000 claims 2
- 238000012552 review Methods 0.000 claims 2
- 230000006837 decompression Effects 0.000 description 6
- 230000008901 benefit Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000004088 simulation Methods 0.000 description 5
- 238000013144 data compression Methods 0.000 description 4
- 230000003993 interaction Effects 0.000 description 3
- 238000013079 data visualisation Methods 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 208000017667 Chronic Disease Diseases 0.000 description 1
- 101001121408 Homo sapiens L-amino-acid oxidase Proteins 0.000 description 1
- 101000827703 Homo sapiens Polyphosphoinositide phosphatase Proteins 0.000 description 1
- 206010020772 Hypertension Diseases 0.000 description 1
- 102100026388 L-amino-acid oxidase Human genes 0.000 description 1
- 102100023591 Polyphosphoinositide phosphatase Human genes 0.000 description 1
- 206010037660 Pyrexia Diseases 0.000 description 1
- 101100012902 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) FIG2 gene Proteins 0.000 description 1
- 101100233916 Saccharomyces cerevisiae (strain ATCC 204508 / S288c) KAR5 gene Proteins 0.000 description 1
- 238000004873 anchoring Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005802 health problem Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000007903 penetration ability Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 150000003839 salts Chemical class 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 208000019116 sleep disease Diseases 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 230000003442 weekly effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
Definitions
- the present invention relates to the field of ship safety management, and in particular to a large ship safety supervision system.
- the present invention proposes a large ship safety supervision system.
- the system is applied to the complex environment of large ships, monitors the real-time position of crew members in the cabin, and conducts comprehensive prevention and management of ship safety and crew health.
- a data transmission method for ship-shore is implemented by a hybrid adaptive compression technology based on model classification and a data transmission link intelligent optimization technology based on fuzzy neural network.
- the hybrid adaptive compression technology based on model classification is used to compress data: first, a multi-classifier is established for classification, and the input layer and output layer of the classification model are as follows:
- Input layer ⁇ data fluctuation frequency, data type ⁇
- the data fluctuation frequency of the input layer refers to the fluctuation frequency of the collected data
- the data type refers to ship data, crew data, hydrological data, meteorological data, and sea condition data
- the classification model automatically selects the output based on the input, which is specifically manifested as:
- the compression method is automatically selected based on the output results ⁇ 0,1,2,3 ⁇ of the model.
- the compression method used is: dynamically set the time window size W t , use the time window to split the data, obtain N groups of data, retain the first data point of the first N-1 groups of data, and the last data point of the Nth group of data, to achieve efficient compression;
- Step 31 read a character from the collected data input stream and proceed to step 32;
- Step 32 If the current code is in the dictionary, use the first character of the current code as the suffix of the current string. If the current string is not in the dictionary, add it to the dictionary and then add the current string to the dictionary. The code is used as a prefix of the string, and then the process goes to step 34;
- Step 33 If the current code is not in the dictionary, take the first character of the prefix as the suffix, add the string to the dictionary, use the current string code as the prefix, and go to step 34;
- Step 34 put the prefix into the output stream and go to step 31;
- This method is an improved spinning door transformation (SDT) compression method
- the traditional SDT algorithm not only has a high compression rate, but also has a short compression and decompression time. However, when the compressed data has a large measurement error or noise, the data compression rate obtained by the SDT algorithm is greatly reduced.
- the present invention creates a mean-based spinning door compression algorithm (MSDT for short) based on mean square error. While inheriting the advantages of the SDT algorithm, the MSDT algorithm reduces the impact of measurement error on compression. MSDT uses mean square error as the accuracy indicator in the algorithm. The algorithm considers the constraint of the total decompression error, rather than the decompression error of each test value in the constraint. The algorithm specifically includes the following seven steps:
- Step 41 (t 0 , y 0 ) is the last stored point, ⁇ is the compression deviation parameter, k upper and k lower are the upper and lower limits of the slope between the last stored point and the new receiving test point, respectively.
- Step 42 When receiving a new measurement point (t, y), calculate
- Step 43 If k lower ⁇ k ⁇ k upper , go to step 45 to continue; otherwise, go to the next step;
- Step 45 Iteratively calculate a, b, c:
- Step 46 If b 2 -4ac ⁇ 0, the point fails the compression test, is stored and used as the new last storage point, and then goes to step 41 to continue compressing data;
- Step 47 Calculate the upper limit k upper and the lower limit k lower :
- step 42 to continue compressing the data.
- the data transmission link intelligent optimization technology based on fuzzy neural network can realize the automatic switching of mobile communication links and satellite broadband links at both ends of the ship and the shore.
- the data transmission link intelligent optimization technology based on fuzzy neural network calculates and selects the optimal link for transmission according to the current network parameters to adapt to different construction environments, which can reduce costs as much as possible while increasing transmission stability.
- the initial input of the data transmission link intelligent optimization technology based on fuzzy neural network is ⁇ RSS, bandwidth, load, delay, signal-to-noise ratio, communication cost of the network link ⁇ , and the final output is ⁇ mobile communication, satellite broadband ⁇ .
- the data transmission link intelligent optimization technology based on fuzzy neural network includes six layers of design, namely input layer, fuzzification layer, fuzzy rule layer, fuzzy decision layer, defuzzification layer and output layer, specifically:
- the input layer sets six input parameters, which are network link attribute parameters, namely received signal strength RSS, network link bandwidth, load, delay, signal-to-noise ratio, and communication cost;
- Fuzzification layer Fuzzify the network link attribute parameters received from the input layer according to the membership function. Obtain fuzzy sets, the received signal strength RSS, bandwidth, load, delay, signal-to-noise ratio and communication cost of the network link are mapped into fuzzy sets of ⁇ low, medium, high ⁇ ;
- Fuzzy rule layer used to calculate the quality of candidate links, and use neural networks to classify candidate link levels into six categories: ⁇ excellent, good, acceptable, bad, poor, and extremely poor ⁇ ;
- i,m 1,2,....,6, X ⁇ R6 , Y ⁇ R6 ⁇ containing N initial training samples, the training process formula for the neural network is designed as follows:
- xi represents the input value
- L is the number of hidden layer neurons
- wj represents the weight of the input layer neuron connected to the jth hidden layer neuron
- Bj represents the weight of the jth hidden layer neuron connected to the output layer neuron
- gj () represents the activation function
- bj represents the hidden layer bias
- N represents the number of samples
- yi represents the output value
- the input layer vector and output layer vector of the neural network classification are:
- Input layer ⁇ RSS, bandwidth, load, delay, signal-to-noise ratio, communication cost ⁇
- Output layer ⁇ excellent, good, acceptable, bad, poor, very poor ⁇
- Defuzzification layer The fuzzy decision layer defines the natural language variables of the candidate link levels as ⁇ excellent, good, acceptable, bad, poor, extremely poor ⁇ , and the defuzzification layer converts the natural language fuzzy sets into precise values for subsequent comparisons.
- ⁇ (x) represents the membership function
- c and ⁇ represent the center and standard deviation of the membership function respectively.
- S l represents the final score of the link;
- Output layer After obtaining the score values of each link, it compares them, selects the highest value in S l , and automatically switches to the link with the highest score (mobile communication link or satellite broadband link) for data transmission.
- the data transmission link intelligent optimization technology based on fuzzy neural network selects the candidate link with the maximum link score value, improves the accuracy of link selection results while reducing calculation overhead, so that the system can always automatically calculate and select the optimal link for transmission according to the current network parameters in real time to adapt to different construction environments.
- a large-scale ship safety supervision system is used to realize crew supervision and ship safety supervision.
- Crew supervision includes real-time cabin position monitoring of crew members and crew health data monitoring.
- Ship safety supervision includes ship sea condition early warning alarm, ship equipment operation status monitoring, ship navigation/construction whole process supervision, and ship remote guidance.
- the present invention has the following advantages and beneficial effects:
- the present invention creates a hybrid adaptive compression technology based on model classification and a data transmission link intelligent optimization technology based on fuzzy neural network to achieve efficient and reliable data transmission between ship and shore, while realizing real-time safety monitoring and management of large ships and analysis and guidance of remote operations of ships. It conducts real-time monitoring and management of the position and health and safety status of crew members in multi-layer cabins of large ships, and can provide real-time position monitoring, trajectory tracking, body temperature monitoring, sleep monitoring, exercise monitoring, blood pressure monitoring and other health parameter monitoring in multi-layer cabins of ship personnel, and has one-button alarm, safety electronic fence, accidental entry into dangerous areas and other alarm functions.
- FIG1 is a system function diagram of a large ship safety supervision system of the present invention.
- FIG2 is a signal flow chart of a large ship safety supervision system according to the present invention.
- FIG3 is a system composition diagram of a data transmission link intelligent optimization algorithm based on a fuzzy neural network
- FIG4 is a diagram showing the classification results of the embodiment of classifying the candidate links into different levels based on a neural network
- Figure 5 is a diagram of the neural network structure
- FIG6 is a comparison diagram of attributes before and after using the hybrid adaptive compression technology based on model classification
- Figure 7 is a large ship virtual reality remote interaction platform
- FIG8 is an example of a three-dimensional reproduction of the ship status of a large ship.
- marine ship data has its own particularities: (1) the signal byte length is fixed; (2) the signal often changes continuously at a certain difference; (3) the signal has tolerance and can be lossily compressed; (4) the signal is generated in real time, with a short sequence, large volume, and carries time-latitude characteristics.
- the signal is generated in real time, with a short sequence, large volume, and carries time-latitude characteristics.
- it is necessary to compress the historical data while striving for extremely high compression speed and high compression ratio.
- the present invention has developed a hybrid adaptive compression technology based on model classification, which has small computational complexity and can track process trend changes for rolling compression, with a compression ratio of up to 24:1.
- the present invention develops a data transmission link intelligent optimization technology based on fuzzy neural network to achieve efficient and reliable data transmission between ship and shore. It uses the link parallel transmission strategy to automatically select the optimal network to transmit data according to the current network status. When the network status changes, the better quality network among multiple networks is selected through the automatic switching mechanism to adapt to different communication environments, while minimizing costs and increasing transmission stability.
- the data transmission link intelligent optimization technology based on fuzzy neural network is further used to transmit data between the ship and the shore.
- the whole scheme has overcome the special environment of marine ships to complete the efficient, reliable and low-cost transmission of data between the ship and the shore.
- a large ship safety supervision system can be used to realize crew supervision and ship safety supervision.
- Crew supervision includes real-time cabin position monitoring of crew members and crew health data monitoring
- ship safety supervision includes ship sea condition warning alarm, ship equipment operation status monitoring, ship navigation/construction full process supervision, and ship remote guidance.
- the software and hardware modules of the system include: Bluetooth beacon M1, wearable terminal device M2 for crew location management and crew health management, Bluetooth gateway M3, infrared thermal imaging face recognition temperature measurement device M4, data dynamic analysis module M5, warning module M6, manual audit terminal M7, large ship safety supervision system terminal M8, real-time guidance module M9 and ship information database M10.
- the data dynamic analysis module M5, warning module M6, and manual audit terminal M7 run on the ship-side server
- the large ship safety supervision system terminal M8 and real-time guidance module M9 run on the shore-side server
- the ship information database M10 runs on the cloud server.
- the signal transmission between the above modules is shown in Figure 2.
- Bluetooth Beacon M1 The Bluetooth Beacon M1 is used in conjunction with the wearable terminal device M2 that integrates crew location management and crew health management to achieve real-time positioning of crew members in multi-layer cabins of large ships.
- the Bluetooth beacon designed for large ships has low power consumption and supports an ultra-long battery life of more than four years. It is suitable for indoor and outdoor environments of large ships. Bluetooth beacons are deployed in specific areas that need to be positioned (crew rooms, work decks, restaurants, etc.) and are fixed in a position away from corners and obstacles. To ensure positioning accuracy, the horizontal spacing of each Bluetooth beacon is evenly controlled at 2m.
- Wearable terminal device M2 for crew position management and crew health management Wearable terminal device M2 for crew position management and crew health management is worn on each crew member to obtain crew health monitoring information and real-time location, and transmit vital sign data back to data dynamic analysis module M5 through Bluetooth gateway M3. Wearable terminal device M2 for crew position management and crew health management includes smart bracelets, smart vests, head-mounted devices, etc. When worn by crew members, it does not affect their daily work and life, and can be used normally in ship swaying and strong noise environments. The device has technical characteristics such as strong penetration ability and long transmission distance.
- the signal can pass through the heavy steel plates of multi-layer cabins of large ships, and can monitor the crew's location information, heart rate information, body temperature information, blood pressure information, blood oxygen information, exercise information, sleep information and other health information in real time.
- the crew's health data and positioning data collected by the device are transmitted to the data dynamic analysis module M5 in the form of a wireless terminal.
- Bluetooth Gateway M3 It is divided into indoor and outdoor Bluetooth gateways for large ships. Specifically:
- the Bluetooth gateway in the ship's cabin adopts an embedded whole-machine design, with built-in high-performance WiFi module, Ethernet module and Bluetooth transceiver module, which can penetrate the steel plates in large cabins to meet data transmission needs.
- the Bluetooth gateway designed for outdoor use on ships is dustproof, waterproof, lightning-proof, and durable. It is suitable for outdoor scenes on specific large ships and supports POE power supply.
- the Bluetooth gateway forms a scanning matrix and transmits the scanned Bluetooth tag signal to the data dynamic analysis module M5 of the ship-side server through POE power supply or connection to the switch.
- Infrared thermal imaging face recognition temperature measurement equipment M4 Use non-contact full-frame human body temperature measurement to measure the human body surface temperature over a long distance and contactlessly, use high-definition pixel network cameras to collect high-definition face images, and use dynamic face recognition algorithms to intelligently identify the crew's identity and name, and achieve dynamic, high-speed, multi-face detection, capture and recognition.
- the equipment automatically generates real-time dynamic thermal images, and uses thermal images to record the passing crew and body temperature in real time, so as to achieve long-distance, large-area detection of crew members in complex ship environments, and rapid temperature screening of multiple people at the same time, with fast temperature measurement, high accuracy and full coverage.
- Data dynamic analysis module M5 The data dynamic analysis module performs statistical analysis on crew data and ship data.
- Crew data includes the physical health data and real-time cabin position data of ship personnel, and performs dynamic statistical analysis on crew data monitored at different time periods.
- the system ensures the safe location and physical health of the crew by real-time dynamic monitoring of the crew.
- the data dynamic analysis module can automatically generate weekly or monthly reports on the physical health data of the crew and feedback to the ship personnel on a regular basis.
- Ship data is automatically collected by sensors and transmitted to the data dynamic analysis module in real time by wireless transmission.
- Ship data includes data on the entire process of ship navigation/construction, data on the production status and equipment operation status of specific ships, and information on sea conditions around the ship.
- the data dynamic analysis module centrally monitors and analyzes the construction status, equipment operation status, maritime alarms and marine environmental safety factors of ships in different countries and regions.
- Warning module M6 The warning module provides real-time alarms for ships and crew members based on the data analysis of the data dynamic analysis module M5.
- Ship alarms involve real-time safety monitoring, safety inspections and safety management of remote ships, including providing real-time ship early warning alarms (such as excessive ship tilt, abnormal ship draft, abnormal operation of key equipment, etc.) and real-time maritime warnings (typhoon, heavy rain, strong wind news, etc.).
- Crew alarms include active alarms and passive alarms. Active alarms refer to crew members actively triggering the SOS button of a wearable terminal device that integrates crew position management and crew health management, actively issuing an alarm and displaying the alarm label, personnel and real-time cabin position in real time on the terminal.
- Passive alarms refer to abnormal alarms for health data such as body temperature, sleep, exercise, blood pressure, heart rate, blood oxygen, etc., as well as alarms for crew members entering/leaving/staying in non-visit areas or specific areas, and mistakenly entering dangerous areas.
- Manual audit terminal M7 The manual audit terminal conducts manual audits on abnormal data. For example, the health data area of a single or a whole group of crew members is entered in the manual audit terminal. When the monitoring value exceeds the preset threshold, an alarm is issued to the warning module. Through manual verification by management personnel or sending personnel on board for inspection, the abnormal data monitored in the data dynamic analysis module is further verified to ensure the rationality and authenticity of the monitoring data.
- Large ship safety supervision system terminal M8 The large ship safety supervision system terminal includes a ship management subsystem and a crew management subsystem, which provide shore managers with the ability to view and manage ship and crew information.
- the ship management subsystem realizes the safety management functions of recording, evidence collection, alarming, etc. of ship safety production activities.
- the ship data at the construction site is published through the Internet using multi-dimensional information fusion technology, HTML5 and WebGL (Web Graphics Library) technology, realizing comprehensive monitoring of marine environmental conditions, ship status, special operation production status, ship equipment location and operation status.
- HTML5 and WebGL Web Graphics Library
- it provides simple data information sharing services, realizes remote construction monitoring, ship engine operation monitoring and fault diagnosis, can adjust the operation plan in time, and realize remote decision-making.
- the crew management subsystem manages the location of large ship operators by equipping them with wearable terminal devices, monitors and manages the indoor location and trajectory of crew members in real time, and has one-click alarm, safety electronic fence, and alarm for accidentally entering dangerous areas.
- electronic fences By setting up electronic fences and SOS one-click emergency alarms in the background management, electronic fences are set for non-access areas or specific areas.
- the system When entering/leaving/staying, the system immediately issues an alarm to ensure the safety of personnel and areas.
- the system also manages and maintains the basic information of large ship operators, manages health information, and provides temperature monitoring, sleep monitoring, exercise monitoring, etc. Monitoring, blood pressure monitoring and other health parameter monitoring services, with temperature screening and alarms, assist in crew management and epidemic tracing during the epidemic.
- the large-scale ship safety monitoring system collects massive data such as ship surrounding environment information, construction status information of major construction equipment, and personnel health information online in real time.
- the present invention compresses the data based on the hybrid adaptive compression technology of model classification.
- a multi-classifier is established for classification.
- the input layer and output layer of the classification model are as follows:
- Input layer ⁇ data fluctuation frequency, data type ⁇
- the data fluctuation frequency of the input layer refers to the fluctuation frequency of the collected data
- the data type refers to ship data, crew data, hydrological data, meteorological data, sea condition data, etc.
- the classification model automatically selects the output based on the input, which is specifically manifested as:
- the compression method is automatically selected based on the output results of the model ⁇ 0,1,2,3 ⁇ , specifically:
- the compression method used is: dynamically set the time window size W t , use the time window to split the data, obtain N groups of data, retain the first data point of the first N-1 groups of data, and the last data point of the Nth group of data, to achieve efficient compression.
- Step 31 Read a character in the collected data input stream and proceed to step 32.
- Step 32 If the current code is in the dictionary, the first character of the current code is used as the suffix of the current string. If the current string is not in the dictionary, it is added to the dictionary, and then the current code is used as the prefix of the string, and go to step 34.
- Step 33 If the current code is not in the dictionary, take the first character of the prefix as the suffix, add the string to the dictionary, use the current string code as the prefix, and go to step 34.
- Step 34 Put the prefix into the output stream and go to step 31.
- S4 When the output is 3, an improved revolving door (Spinning Door Transformation, SDT) compression method is used.
- the traditional SDT algorithm not only has a high compression rate, but also has a short compression and decompression time. However, when the compressed data has a large measurement error or noise, the data compression rate obtained by the SDT algorithm is greatly reduced.
- the present invention creates a mean-based spinning door compression algorithm (MSDT) based on mean square error. While inheriting the advantages of the SDT algorithm, the MSDT algorithm reduces the impact of measurement error on compression. MSDT uses mean square error as the accuracy indicator in the algorithm.
- the algorithm considers the constraint of the total decompression error, rather than the decompression error of each test value in the constraint. The algorithm specifically includes the following seven steps:
- Step 41 (t 0 , y 0 ) is the last stored point, ⁇ is the compression deviation parameter, k upper and k lower are the upper and lower limits of the slope between the last stored point and the new receiving test point, respectively.
- Step 42 When receiving a new measurement point (t, y), calculate
- Step 43 If k lower ⁇ k ⁇ k upper , go to step 45 to continue; otherwise, go to the next step;
- Step 45 Iteratively calculate a, b, c:
- Step 46 If b 2 -4ac ⁇ 0, the point fails the compression test, is stored and used as the new last storage point, and then goes to step 41 to continue compressing data;
- Step 47 Calculate the upper limit k upper and the lower limit k lower :
- step 42 to continue compressing the data.
- the present invention firstly adopts a hybrid adaptive compression technology based on model classification.
- a hybrid adaptive compression technology based on model classification.
- the algorithm has a small amount of computation and can track the trend changes of the process.
- the compressed data is transmitted using the data transmission link intelligent optimization technology based on fuzzy neural network.
- the hybrid adaptive compression technology based on model classification can achieve a compression ratio of 24:1, and the compression efficiency is improved by more than 18%.
- the data After the data is compressed and stored using a hybrid adaptive compression technology based on model classification, it is transmitted using a data transmission link intelligent optimization technology based on a fuzzy neural network.
- the present invention develops a data transmission link intelligent optimization technology based on fuzzy neural network to realize the automatic switching of mobile communication links and satellite broadband links at both ends of the ship and the shore.
- Data transmission based on fuzzy neural network The intelligent link optimization technology calculates and selects the optimal link for transmission based on the current network parameters to adapt to different construction environments. It can minimize costs while increasing transmission stability.
- the initial input of the data transmission link intelligent optimization technology based on fuzzy neural network is ⁇ RSS, bandwidth, load, delay, signal-to-noise ratio, communication cost of the network link ⁇ , and the final output is ⁇ mobile communication, satellite broadband ⁇ .
- the data transmission link intelligent optimization technology based on fuzzy neural network includes six layers of design, namely input layer, fuzzification layer, fuzzy rule layer, fuzzy decision layer, defuzzification layer and output layer, specifically:
- the input layer sets six input parameters, which are network link attribute parameters, namely received signal strength RSS, network link bandwidth, load, delay, signal-to-noise ratio, and communication cost;
- Fuzzification layer The network link attribute parameters received from the input layer are fuzzified according to the membership function to obtain a fuzzy set.
- the received signal strength RSS, bandwidth, load, delay, signal-to-noise ratio and communication cost of the network link are mapped into a fuzzy set of ⁇ low, medium, high ⁇ ;
- Fuzzy rule layer used to calculate the quality of candidate links, and use neural networks to classify candidate link levels into six categories: ⁇ excellent, good, acceptable, bad, poor, and extremely poor ⁇ ;
- i,m 1,2,....,6, X ⁇ R6 , Y ⁇ R6 ⁇ containing N initial training samples, the training process formula for the neural network is designed as follows:
- xi represents the input value
- L is the number of hidden layer neurons
- wj represents the weight of the input layer neuron connected to the jth hidden layer neuron
- Bj represents the weight of the jth hidden layer neuron connected to the output layer neuron
- gj () represents the activation function
- bj represents the hidden layer bias
- N represents the number of samples
- yi represents the output value
- the input layer vector and output layer vector of the neural network classification are:
- Input layer ⁇ RSS, bandwidth, load, delay, signal-to-noise ratio, communication cost ⁇
- Output layer ⁇ excellent, good, acceptable, bad, poor, very poor ⁇
- Defuzzification layer The fuzzy decision layer defines the natural language variables of the candidate link levels as ⁇ excellent, good, acceptable, bad, poor, extremely poor ⁇ , and the defuzzification layer converts the natural language fuzzy sets into precise values for subsequent comparisons.
- ⁇ (x) represents the membership function
- c and ⁇ represent the center and standard deviation of the membership function respectively.
- S l represents the final score of the link;
- Output layer After obtaining the score values of each link, it compares them, selects the highest value in S l , and automatically switches to the link with the highest score (mobile communication link or satellite broadband link) for data transmission.
- the data transmission link intelligent optimization technology based on fuzzy neural network selects the candidate link with the maximum link score value, improves the accuracy of link selection results while reducing calculation overhead, so that the system can always automatically calculate and select the optimal link for transmission according to the current network parameters in real time to adapt to different construction environments.
- Real-time guidance module M9 The real-time guidance module provides visual analysis and construction guidance for remote operations of large ships, and reproduces the status of large ships in three dimensions based on virtual reality technology. Through the data dynamic analysis module M5, real-time statistical analysis of the safety status, production status and equipment operation status of the ship can be performed. The construction process parameters of large ships can be viewed and managed remotely in real time, and remote construction guidance and suggestions can be provided based on the dynamic comprehensive status of the actual project. A digital twin mapping model of real ship data and simulation data is established, and a semi-physical simulation three-dimensional remote interactive construction platform is used to realize comprehensive numerical simulation and ship scene reproduction of the entire process of navigation/production of large ships.
- the remote control of the digital twin of the ship breaks through the limitations of time and space, allowing experts to guide on-site ship operations on the shore.
- Experts view the visual analysis and playback results of historical data, and provide experts with gestures, pictures, and text.
- the interactive guidance function with features such as the following: Through the visual construction guidance of senior technicians, it can effectively solve urgent, difficult and dangerous problems on site, and strongly support major engineering projects of large ships.
- Figure 7 is a large ship virtual reality remote interaction platform, which realizes comprehensive numerical simulation of the entire process of large ship navigation/production and ship scene reproduction through a semi-physical simulation construction three-dimensional remote interaction platform.
- the ship's condition can be reproduced in three dimensions by using technologies such as twin model and real physical space operation data visualization, remote video fusion, and real-time image fusion, as shown in Figure 8 as an example.
- Ship information database M10 A large amount of ship information, physical data of crew members in different conditions and real-time location data are stored in the information database to realize the historical data backtracking of ships and their personnel.
- the system establishes independent files for each ship and each crew member.
- the ship-side server pushes data to the shore-side server, thereby synchronizing the data of the server on the ship and viewing the data.
- the local server (shore-side server) pushes synchronously to the cloud server to realize the synchronization of ship, shore and cloud information.
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
一种大型船舶安全监管系统,用于实现船员监管和船舶安全监管,其中船员监管包括船员多层舱内实时位置监控和船员健康数据监测,船舶安全监管包括船舶海况预警警报、船舶设备运行状况监控、船舶航行/施工全过程监管及船舶远程指导。基于模型分类的混合自适应压缩技术和模糊神经网络的数据传输链路智能优选技术进行船、岸间的数据传输,同时实现大型船舶实时安全监测管理和船舶远程作业分析指导,针对大型船舶多层舱内的船员位置和健康安全状态进行实时监控管理,可提供船舶人员多层舱内的实时位置监控、轨迹跟踪、体温监测、睡眠监测、运动监测、血压监测等健康参数监测,并具备一键报警、安全电子围栏、误闯危险区域等报警功能。
Description
本发明涉及船舶安全管理领域,特别涉及一种大型船舶安全监管系统。
大型工程船舶施工现场的地质、水文和气候环境复杂,面临风浪流等多项水域自然环境因素的动态变化影响,安全风险突出,传统的船舶安全监管技术方法的实时性、稳定性和可靠性难以保证,大型船舶的安全监控管理存在短板。大型船舶结构普遍为钢制结构,船体坚硬,船舶空间相对封闭且狭小,传统的定位系统无法有效实时追踪大型船舶多层舱内船舶人员的实时位置。船员在施工作业中,存在通讯不畅、缺乏精确的实时舱内定位和一键电子报警功能,缺少安全电子围栏的设置和误闯危险区域的警报。船员长期处于“四高两缺一多”(高温、高湿、高盐、高辐射和缺土、缺水、多突风)的恶劣环境,工作任务繁重、环境单调,其身体健康状况容易受不同程度的影响,可能存在血压上升、睡眠障碍、体温不稳、心率不齐等各种健康问题。船上环境复杂,一般的陆用健康监测系统难以在高温、高湿且剧烈摇晃的大型施工船舶中有效使用,在长期航行过程中,船员可以获取的医疗资源有限,可能导致船员在发生慢性病变的中间过程中,错失了最佳治疗时间。
发明内容
为了解决上述问题,本发明提出了一种大型船舶安全监管系统。该系统应用于大型船舶复杂环境,针对船员实时舱内位置进行监控,并对船舶安全和船员健康进行全面防范管理。
本发明技术方案如下:
技术方案一
一种应用于船岸间的数据传输方法,先后基于模型分类的混合自适应压缩技术和基于模糊神经网络的数据传输链路智能优选技术来实现。
所述基于模型分类的混合自适应压缩技术,用于对数据进行压缩:首先建立多分类器进行分类,分类模型的输入层、输出层如下:
输入层={数据波动频率,数据类型}
输出层={0,1,2,3}
其中,输入层的数据波动频率指采集到数据的波动频率,数据类型是指船舶数据、船员数据、水文数据、气象数据、海况数据;
分类模型根据输入自动选择输出,具体表现为:
S1:适用于船舶抛锚停工场景下,数据波动频率无波动时,输出结果倾向为0;
S2:数据波动频率缓慢(数值在100秒及以上时间只有±3%以内波动)、数据类型为船舶数据、水文数据时,输出结果倾向为1;
S3:数据波动频率较慢(数值在30秒及以上时间无变化),且数据类型为气象数据、海况数据时,输出结果倾向为2;
S4:针对除S1、S2或S3的其他大部分的情形,倾向输出结果为3;
根据模型的输出结果{0,1,2,3}自动选择压缩方法。
具体为:
S1:当输出为0时,采用的压缩方法:动态设置时间窗口大小Wt,利用时间窗口分割数据,得到N组数据,保留前N-1组数据的第一个数据点,和第N组数据的最后一个数据点,实现高效压缩;
S2:当输出为1时,采用的压缩方法:将一个相同值的连续串用其值和串长(重复的个数)的数对二元组来替代,解压缩时根据字符及连续相同字符的个数可恢复至原来的数据;
S3:当输出为2时,采用的压缩方法为以下四个步骤:
步骤31:在采集的数据输入流中读取一个字符,进入步骤32;
步骤32:如果当前编码在字典中,则把当前编码的第一个字符作为当前字符串的后缀,如果当前字符串不在字典中,就把它加入到字典中,然后把当前
编码作为字符串的前缀,转到步骤34;
步骤33:如果当前编码不在字典中,就把前缀的第一个字符作为后缀,把字符串加入到字典中,用当前串的编码作前缀,转到步骤34;
步骤34:把前缀放到输出流,转到步骤31;
S4:当输出为3时,采用基于均方误差的旋转门压缩算法(Mean-based Spinning Door Transformation,简称MSDT);
该方法为改进的旋转门(Spinning Door Transformation,简称SDT)压缩方法;
传统的SDT算法不仅压缩率高,而且压缩和解压时间都很短,但是当被压缩的数据具有很大的测量误差或噪声时,SDT算法得到的数据压缩率大大下降。为减少测量噪声对数据压缩的影响,本发明开创一种基于均方误差的旋转门压缩算法(Mean-based Spinning Door Transformation,简称MSDT)。MSDT算法在继承SDT算法优势的同时,降低了测量误差对压缩的影响。MSDT运用均方误差作为算法中的精度指标,算法考虑总的解压缩误差的约束,而非约束中每个测试值的解压缩误差,算法具体包含以下七个步骤:
步骤41:(t0,y0)是上一个存储点,Δ是压缩偏差参数,kupper和klower分别是上一个存储点和新接收测试点之间斜率的上、下限。
初始化a=0,b=0,c=0,kupper=+∞,klower=-∞;
步骤42:当接收到新测点(t,y),计算
步骤43:如果klower<k<kupper,转到步骤45继续,否则,执行下一步;
步骤44:该点未通过压缩测试,被存储并被用作新的上一个存储点,然后重新初始化三个参数a,b,c:
a=τ2
b=-2kτ2
c=k2τ2-Δ2
a=τ2
b=-2kτ2
c=k2τ2-Δ2
这里,τ和k的值利用新的上一个存储点进行计算,然后转到步骤46继续;
步骤45:迭代计算a,b,c:
步骤46:如果b2-4ac≤0,该点未通过压缩测试,被存储并将其作为新的上一个存储点,然后转到步骤41继续压缩数据;
步骤47:计算上限kupper和下限klower:
然后,转到步骤42继续压缩数据。
所述基于模糊神经网络的数据传输链路智能优选技术,实现船、岸两端移动通信链路和卫星宽带链路的自动切换。基于模糊神经网络的数据传输链路智能优选技术根据当前网络参数计算并选择最优链路进行传输,以适应不同的施工环境,能在尽量降低成本的同时增加传输稳定性。
基于模糊神经网络的数据传输链路智能优选技术的初始输入为{网络链路的接收信号强度RSS、带宽、负载、延时、信噪比、通信费用},最终输出为{移动通信,卫星宽带}。基于模糊神经网络的数据传输链路智能优选技术包含六层设计,分别为输入层、模糊化层、模糊规则层、模糊决策层、解模糊层和输出层,具体为:
输入层:输入层设置六个输入参数,所述六个输入参数为网络链路属性参数,分别为接收信号强度RSS、网络链路带宽、负载、延时、信噪比、通信费用;
模糊化层:根据隶属度函数将从输入层接收到的网络链路属性参数模糊化,
得到模糊集合,网络链路的接收信号强度RSS、带宽、负载、延时、信噪比和通信费用被映射成为{低,中,高}的模糊集合;
模糊规则层:用来计算候选链路的优劣等级,利用神经网络将候选链路等级划分为{极好,较好,可以,不好,较差,极差}六大类;
模糊决策层:利用神经网络分类器搭建六输入、六输出的神经网络架构。给定含N个初始训练样本的数据集{X,Y}={(xi,ym)|i,m=1,2,....,6,X∈R6,Y∈R6},设计神经网络的训练过程公式如下:
其中xi代表输入值,L为隐藏层神经元个数,wj代表输入层神经元连接第j个隐藏层神经元的权值,Bj代表第j个隐藏层神经元连接输出层神经元的权值,gj()代表激活函数,bj代表隐藏层偏置,N代表样本个数,yi代表输出值;
神经网络分类的输入层向量和输出层向量分别为:
输入层={RSS、带宽、负载、延时、信噪比、通信费用}
输出层={极好,较好,可以,不好,较差,极差}
解模糊层:模糊决策层将候选链路等级的自然语言变量定义为{极好,较好,可以,不好,较差,极差},解模糊层将自然语言模糊集合转化为精确数值用于后续的比较。
解模糊化的计算公式如下:
其中p取值在[1,2],μ(x)表示隶属度函数,c和σ分别为隶属函数的中心和标准差。Sl表示链路最终得分;
输出层:在获得的各个链路的得分值之后进行比较,选取Sl中的最高值,并自动切换到得分最高的链路(移动通信链路或卫星宽带链路)进行数据传输。
基于模糊神经网络的数据传输链路智能优选技术选取具有最大链路得分值的候选链路,在减少计算开销的情况下提高链路选择结果的准确性,使得系统能够始终实时地自动根据当前网络参数计算并选择最优链路进行传输,以适应不同的施工环境。
技术方案二
一种大型船舶安全监管系统,用于实现船员监管和船舶安全监管,其中,船员监管包括船员实时舱内位置监控、船员健康数据监测,船舶安全监管包括船舶海况预警警报、船舶设备运行状况监控、船舶航行/施工全过程监管、船舶远程指导。
与现有技术相比,本发明具有如下优点和有益效果:
本发明创造基于模型分类的混合自适应压缩技术和基于模糊神经网络的数据传输链路智能优选技术实现船、岸间数据高效可靠传输,同时实现大型船舶实时安全监测管理和船舶远程作业分析指导,针对大型船舶多层舱内的船员位置和健康安全状态进行实时监控管理,可提供船舶人员多层舱内的实时位置监控、轨迹跟踪、体温监测、睡眠监测、运动监测、血压监测等健康参数监测,并具备一键报警、安全电子围栏、误闯危险区域等报警功能。
图1为本发明一种大型船舶安全监管系统的系统功能图;
图2为本发明一种大型船舶安全监管系统的信号流程图;
图3为基于模糊神经网络的数据传输链路智能优选算法的系统组成图;
图4为实施例基于神经网络将候选链路等级划分的分类结果图;
图5为神经网络结构图;
图6为采用基于模型分类的混合自适应压缩技术前后的属性对比图;
图7为大型船舶虚拟现实远程交互平台;
图8为大型船舶实船的三维重现船舶状况示例。
有别于传统商业数据,海上船舶数据存在其特殊性:(1)信号字节长度固定;(2)信号常按一定差值连续变化;(3)信号具备容差性,能够对其有损压缩;(4)信号实时产生,序列短,体量大,携带时间纬度特征。为了有效使用船载磁盘存储和减少船岸两端传输的数据量,需要对历史数据进行压缩,同时力求极高的压缩速度及高压缩比。
为了突破水上施工海量数据的传输受限难题,本发明研发了基于模型分类的混合自适应压缩技术,运算量小,可以跟踪过程趋势变化进行滚动压缩,压缩比高达24:1。
目前大型船舶主要采用卫星宽带或陆地移动基站信号进行数据传输。前者覆盖广,但费用昂贵,且带宽受限;后者覆盖范围主要为距离陆地基站数十公里以内,通常价格低廉,且网络带宽相对较高。在目前普遍使用的依靠单一通信制式进行数据传输的船舶通信系统中,通常会发生通信不稳定的问题。如何在恶劣环境影响下最大程度地保障通信的可靠性,同时保持传输速率是海上数据传输是本领域现阶段存在的痛点、难点。
为此本发明开发基于模糊神经网络的数据传输链路智能优选技术以实现船、岸间数据高效可靠传输。其利用链路并行传输策略,自动根据当前网络状况选择最优网络传输数据,当网络状况发生变化时,通过自动切换机制选择多种网络中较好的优质网络,以适应不同的通信环境,在尽量降低成本的同时增加传输稳定性。
上述基于模型分类的混合自适应压缩技术压缩存储之后,进一步基于模糊神经网络的数据传输链路智能优选技术在船、岸两端进行数据传输。两算法技术先后相配合下,整个方案极好地克服了海上船舶特殊环境下完成数据在船、岸间高效、可靠、低成本的传输。
下面将结合具体实施例及其附图对本申请提供的技术方案作进一步说明。
结合下面说明,本申请的优点和特征将更加清楚。
如图1所示,一种大型船舶安全监管系统,可用于实现船员监管和船舶安全监管。其中,船员监管包括船员实时舱内位置监控、船员健康数据监测,船舶安全监管包括船舶海况预警警报、船舶设备运行状况监控、船舶航行/施工全过程监管、船舶远程指导。
如图2所示,该系统的软硬件模块包括:蓝牙信标M1、集船员位置管理及船员健康管理的可穿戴式终端设备M2、蓝牙网关M3、红外热成像人脸识别测温设备M4、数据动态分析模块M5、警示模块M6、人为审核端M7、大型船舶安全监管系统终端M8、实时指导模块M9和船舶信息数据库M10。所述数据动态分析模块M5、警示模块M6、人为审核端M7运行于船端服务器,所述大型船舶安全监管系统终端M8、实时指导模块M9运行于岸端服务器,所述船舶信息数据库M10,运行于云端服务器。上述各个模块之间的信号传输如图2所示。
蓝牙信标M1:蓝牙信标M1和集船员位置管理及船员健康管理的可穿戴式终端设备M2配合使用,实现大型船舶多层舱内船员实时定位。针对大型船舶设计的蓝牙信标功耗低,支持超过四年的超长续航,适用于大型船舶室内和室外露天环境。蓝牙信标部署在特定需要定位的区域(船员房间、工作甲板、餐厅等),固定在远离墙角和障碍物的位置,为保证定位精度,各个蓝牙信标水平间距均匀控制在2m。
集船员位置管理及船员健康管理的可穿戴式终端设备M2:集船员位置管理及船员健康管理的可穿戴式终端设备M2配戴在每位船员身上,用于获取船员健康监测信息和实时位置,并通过蓝牙网关M3将体征数据回传数据动态分析模块M5。集船员位置管理及船员健康管理的可穿戴式终端设备M2包括智能手环、智能背心、头戴设备等,船员穿戴时不影响日常作业和生活,且能在船舶摇摆和强噪音环境下正常使用。该设备具有穿透能力强、传输距离远等技术特点,信号可穿越大型船舶多层舱的厚重钢板,可以实时监测船员的位置信息、心率信息、体温信息、血压信息、血氧信息、运动信息、睡眠信息等健康
信息。通过该设备收集的船员的身体健康数据和定位数据采用无线终端的模式传输至数据动态分析模块M5。
蓝牙网关M3:分为大型船舶室内和室外蓝牙网关。具体为:
(1)船舶室内的蓝牙网关采用嵌入式整机式设计,内嵌高性能wifi模块、以太网模块以及蓝牙收发模块,可穿透大型船舱内钢板,满足数据传输需求。
(2)船舶室外设计的蓝牙网关具备防尘、防水、防雷、坚固耐用等特点,适用于特定大型船舶室外场景,支持POE供电。通过蓝牙网关组成扫描矩阵,并经POE供电或连接至交换机,将扫描到的蓝牙标签信号传输到船端服务器的数据动态分析模块M5。
红外热成像人脸识别测温设备M4:运用非接触全幅人体测温方式对人体表面温度进行远距离非接触性测温,利用高清像素网络摄像机采集高清人脸图像,并配套使用动态人脸识别算法智能识别船员身份和姓名,实现动态、高速、多人脸检测、抓拍和识别。船员通过时,设备自动生成实时动态热像图,利用热像图实时记录通过船员及体温,实现船舶复杂环境下的船员远距离、大面积检测、多人同时快速体温筛选,测温快、精度高、覆盖全。同时配合大型船舶多层舱内船员实时位置,追踪人员行进轨迹,特别是疑似发热人员的活动轨迹,快速定位,对通行人员、通行数据、体温异常人员数据等全局管控。红外热成像人脸识别测温设备监测的数据传输至数据动态分析模块M5。
数据动态分析模块M5:数据动态分析模块对船员数据和船舶数据进行统计分析。船员数据包含船舶人员的身体健康数据和实时舱内位置数据,针对不同时段监测的船员数据进行动态统计分析。系统通过对船员的动态实时监测,保证船员的安全位置及身体健康。数据动态分析模块可定期自动生成船员身体健康状况数据周报或月报并反馈船舶人员。船舶数据由传感器自动采集并采用无线传输方式实时传送至数据动态分析模块。船舶数据包括船舶航行/施工全过程的数据、特定船舶生产状态和设备运行状态的数据,以及船舶周边海况信息数据。数据动态分析模块将不同的国家和地区船舶的施工状况、设备运行状况、海事警报和海洋环境安全因素进行集中监控分析。
警示模块M6:警示模块根据数据动态分析模块M5的数据分析,提供船舶和船员实时报警。船舶报警涉及远程船舶的实时安全监控、安全检查和安全管理,包含提供实时船舶预警警报(如船舶倾斜超限、船舶吃水异常、关键设备运行异常等现象)和实时海事警告(台风、暴雨、大风消息等)。船员报警包括主动报警和被动报警。主动报警指船员主动触发集船员位置管理及船员健康管理的可穿戴式终端设备的SOS按键,主动发出告警并在终端实时显示告警的标签、人员及实时舱内位置。被动报警指体温、睡眠、运动、血压、心率、血氧等健康数据异常警报,以及船员进入/离开/滞留非访问区域或特定区域,误闯危险区域报警。
人为审核端M7:人为审核端针对异常数据进行人为审核,如在人为审核端输入单个或整批船员的健康数据区域,当监测值超出预设阈值时发出警报至警示模块。通过管理人员人工核查或派人员上船稽查,进一步核实数据动态分析模块中监测异常的数据,保证监测数据的合理性和真实性。大型船舶安全监管系统终端M8:大型船舶安全监管系统终端包括船舶管理子系统和船员管理子系统,提供岸端管理者查看并管理船舶信息、船员信息。
其中船舶管理子系统实现对船舶安全生产活动的记录、取证、报警等安全管理功能。利用多维信息融合技术、HTML5及WebGL(Web Graphics Library)技术将施工现场船舶数据通过互联网发布,实现海洋环境工况、船舶状态、特殊作业生产状态、船舶设备位置及运行状态的全面监控。结合数据走势图、多倍速的历史回放功能,提供简便的数据信息共享服务,实现远程的施工监控、船机运行监控和故障诊断,可及时调整作业方案,实现远程决策。
船员管理子系统通过给大型船舶作业人员配备可穿戴式终端设备来实现对作业人员的位置管理,实时监控管理船员室内位置、轨迹跟踪,具备一键报警、安全电子围栏、误闯危险区域报警功能。通过在后台管理中设置电子围栏和SOS一键紧急报警,对非访问区域或特定区域进行电子围栏设置,进入/离开/滞留时系统立即告警,保障人员及区域安全。系统同时实现对大型船舶作业人员的基本信息的管理维护、健康状况信息管理等,提供体温监测、睡眠监测、运动
监测、血压监测等健康参数监测服务,具备体温筛查及警报,辅助疫情期间的船员管理和疫情追溯。
大型船舶安全监控系统在线实时采集船舶周围环境信息、重大施工装备施工状态信息、人员健康信息等海量数据。本发明基于模型分类的混合自适应压缩技术对数据进行压缩。首先建立多分类器进行分类,分类模型的输入层、输出层如下:
输入层={数据波动频率,数据类型}
输出层={0,1,2,3}
其中,输入层的数据波动频率指采集到数据的波动频率,数据类型是指船舶数据、船员数据、水文数据、气象数据、海况数据等。
分类模型根据输入自动选择输出,具体表现为:
S1:数据波动频率无波动时(此时对应的场景往往为船舶抛锚、停工等),输出结果倾向为0。
S2:数据波动频率缓慢(数值在100秒及以上时间只有±3%以内波动)、数据类型为船舶数据、水文数据时,输出结果倾向为1。
S3:数据波动频率较慢(数值在30秒及以上时间无变化)、数据类型为气象数据、海况数据时,输出结果倾向为2。
S4:针对除S1、S2、S3的其他情形,倾向输出结果为3。分类模型大部分情况输出结果为3。
根据模型的输出结果{0,1,2,3}自动选择压缩方法,具体为:
S1:当输出为0时,采用的压缩方法:动态设置时间窗口大小Wt,利用时间窗口分割数据,得到N组数据,保留前N-1组数据的第一个数据点,和第N组数据的最后一个数据点,实现高效压缩。
S2:当输出为1时,采用的压缩方法:将一个相同值的连续串用其值和串长(重复的个数)的数对二元组来替代,解压缩时根据字符及连续相同字符的个数可恢复至原来的数据。
S3:当输出为2时,采用的压缩方法为以下四个步骤:
步骤31:在采集的数据输入流中读取一个字符,进入步骤32。
步骤32:如果当前编码在字典中,则把当前编码的第一个字符作为当前字符串的后缀,如果当前字符串不在字典中,就把它加入到字典中,然后把当前编码作为字符串的前缀,转到步骤34。
步骤33:如果当前编码不在字典中,就把前缀的第一个字符作为后缀,把字符串加入到字典中,用当前串的编码作前缀,转到步骤34。
步骤34:把前缀放到输出流,转到步骤31。
S4:当输出为3时,采用改进的旋转门(Spinning Door Transformation,简称SDT)压缩方法。传统的SDT算法不仅压缩率高,而且压缩和解压时间都很短,但是当被压缩的数据具有很大的测量误差或噪声时,SDT算法得到的数据压缩率大大下降。为减少测量噪声对数据压缩的影响,本发明开创一种基于均方误差的旋转门压缩算法(Mean-based Spinning Door Transformation,简称MSDT),MSDT算法在继承SDT算法优势的同时,降低了测量误差对压缩的影响。MSDT运用均方误差作为算法中的精度指标,算法考虑总的解压缩误差的约束,而非约束中每个测试值的解压缩误差,算法具体包含以下七个步骤:
步骤41:(t0,y0)是上一个存储点,Δ是压缩偏差参数,kupper和klower分别是上一个存储点和新接收测试点之间斜率的上、下限。
初始化a=0,b=0,c=0,kupper=+∞,klower=-∞;
步骤42:当接收到新测点(t,y),计算
步骤43:如果klower<k<kupper,转到步骤45继续,否则,执行下一步;
步骤44:该点未通过压缩测试,被存储并被用作新的上一个存储点,然后重新初始化三个参数a,b,c:
a=τ2
b=-2kτ2
c=k2τ2-Δ2
a=τ2
b=-2kτ2
c=k2τ2-Δ2
这里,τ和k的值利用新的上一个存储点进行计算,然后转到步骤46继续;
步骤45:迭代计算a,b,c:
步骤46:如果b2-4ac≤0,该点未通过压缩测试,被存储并将其作为新的上一个存储点,然后转到步骤41继续压缩数据;
步骤47:计算上限kupper和下限klower:
然后,转到步骤42继续压缩数据。
为了应对海上特殊通信环境,本发明首先采用基于模型分类的混合自适应压缩技术,一方面经压缩仅用较低的带宽实现大量数据传输的需求,另一方面算法运算量小,且可跟踪过程趋势变化。接着,压缩后的数据采用基于模糊神经网络的数据传输链路智能优选技术进行传输。
如图6所示,经多次实测,基于模型分类的混合自适应压缩技术,压缩比可达24:1,压缩比效率提升18%以上。
数据经基于模型分类的混合自适应压缩技术压缩存储之后,采用基于模糊神经网络的数据传输链路智能优选技术进行传输。
在船、岸两端数据传输过程中,移动通信具有成本低、带宽高的优点,但网络覆盖范围不大;而卫星宽带的覆盖范围广,但成本很高,所以推广应用难。为此本发明开辟基于模糊神经网络的数据传输链路智能优选技术,实现船、岸两端移动通信链路和卫星宽带链路的自动切换。基于模糊神经网络的数据传输
链路智能优选技术根据当前网络参数计算并选择最优链路进行传输,以适应不同的施工环境,能在尽量降低成本的同时增加传输稳定性。
基于模糊神经网络的数据传输链路智能优选技术的初始输入为{网络链路的接收信号强度RSS、带宽、负载、延时、信噪比、通信费用},最终输出为{移动通信,卫星宽带}。基于模糊神经网络的数据传输链路智能优选技术包含六层设计,分别为输入层、模糊化层、模糊规则层、模糊决策层、解模糊层和输出层,具体为:
输入层:输入层设置六个输入参数,所述六个输入参数为网络链路属性参数,分别为接收信号强度RSS、网络链路带宽、负载、延时、信噪比、通信费用;
模糊化层:根据隶属度函数将从输入层接收到的网络链路属性参数模糊化,得到模糊集合,网络链路的接收信号强度RSS、带宽、负载、延时、信噪比和通信费用被映射成为{低,中,高}的模糊集合;
模糊规则层:用来计算候选链路的优劣等级,利用神经网络将候选链路等级划分为{极好,较好,可以,不好,较差,极差}六大类;
模糊决策层:利用神经网络分类器搭建六输入、六输出的神经网络架构(见图5)。给定含N个初始训练样本的数据集{X,Y}={(xi,ym)|i,m=1,2,....,6,X∈R6,Y∈R6},设计神经网络的训练过程公式如下:
其中xi代表输入值,L为隐藏层神经元个数,wj代表输入层神经元连接第j个隐藏层神经元的权值,Bj代表第j个隐藏层神经元连接输出层神经元的权值,gj()代表激活函数,bj代表隐藏层偏置,N代表样本个数,yi代表输出值;
神经网络分类的输入层向量和输出层向量分别为:
输入层={RSS、带宽、负载、延时、信噪比、通信费用}
输出层={极好,较好,可以,不好,较差,极差}
解模糊层:模糊决策层将候选链路等级的自然语言变量定义为{极好,较好,可以,不好,较差,极差},解模糊层将自然语言模糊集合转化为精确数值用于后续的比较。
解模糊化的计算公式如下:
其中p取值在[1,2],μ(x)表示隶属度函数,c和σ分别为隶属函数的中心和标准差。Sl表示链路最终得分;
输出层:在获得的各个链路的得分值之后进行比较,选取Sl中的最高值,并自动切换到得分最高的链路(移动通信链路或卫星宽带链路)进行数据传输。
基于模糊神经网络的数据传输链路智能优选技术选取具有最大链路得分值的候选链路,在减少计算开销的情况下提高链路选择结果的准确性,使得系统能够始终实时地自动根据当前网络参数计算并选择最优链路进行传输,以适应不同的施工环境。
实时指导模块M9:实时指导模块提供大型船舶远程作业的可视化分析和施工指导,并基于虚拟现实技术三维重现大型船舶状况。通过数据动态分析模块M5实时统计分析船舶的安全状态、生产状态和设备的运行状态,可远程实时查看和管理大型船舶施工工艺参数,并根据实际工程等动态综合状态提供远程施工指导和建议。建立实船数据与仿真数据数字孪生映射模型,利用半物理仿真的施工三维远程交互平台实现对大型船舶航行/生产全过程的综合数值模拟仿真和船舶场景复现。船舶数字孪生远程操控突破时间和空间限制,实现专家在岸端即可指导现场船舶作业,基于孪生体模型和真实物理空间的作业数据可视化、远程视频融合和实时影像融合等技术,实现远程指导的“量化传达”,专家查看历史数据的可视化分析和回放结果,提供专家通过手势、图画、文字
等特色的互动指导功能。通过资深技术人员可视化的施工指导有效解决现场急难险重问题,有力支撑大型船舶重大工程项目。
图7为大型船舶虚拟现实远程交互平台,通过半物理仿真的施工三维远程交互平台实现对大型船舶航行/生产全过程的综合数值模拟仿真和船舶场景复现。
利用孪生体模型和真实物理空间的作业数据可视化、远程视频融合和实时影像融合等技术可三维重现船舶状况,作为示例如图8所示。
船舶信息数据库M10:将大量船舶信息、船员不同状况的身体数据和实时位置数据储存至信息数据库内,实现对船舶及其人员的历史数据回溯。系统针对每艘船舶和每位船员分别建立独立的档案。在大型船舶有联网情况下,由船端服务器向岸端服务器推送数据,从而同步船上服务器的数据,并进行数据查看。通过配置HTTP推送路径方式,由本地服务器(岸端服务器)同步推送至云端服务器,实现船、岸、云端信息同步。
上述描述仅是对本申请较佳实施例的描述,并非是对本申请范围的任何限定。任何熟悉该领域的普通技术人员根据上述揭示的技术内容做出的任何变更或修饰均应当视为等同的有效实施例,均属于本申请技术方案保护的范围。
Claims (10)
- 一种大型船舶安全监管系统,其特征在于,用于实现船员监管和船舶安全监管,其中,所述船员监管包括船员实时舱内位置监控、船员健康数据监测,所述船舶安全监管包括船舶海况预警警报、船舶设备运行状况监控、船舶航行/施工全过程监管、船舶远程指导。
- 如权利要求1所述的大型船舶安全监管系统,其特征在于,具体的,该系统包括:蓝牙信标、集船员位置管理及船员健康管理的可穿戴式终端设备、蓝牙网关、红外热成像人脸识别测温设备、数据动态分析模块、警示模块、人为审核端、大型船舶安全监管系统终端、实时指导模块和船舶信息数据库;所述数据动态分析模块、警示模块、人为审核端运行于船端服务器;所述大型船舶安全监管系统终端、实时指导模块运行于岸端服务器;所述船舶信息数据库运行于云端服务器。
- 如权利要求2所述的大型船舶安全监管系统,其特征在于,所述蓝牙信标和集船员位置管理及船员健康管理的可穿戴式终端设备配合使用,实现大型船舶多层舱内船员实时定位;所述蓝牙信标部署在规划所需定位的区域,并固定在远离墙角和障碍物的位置;所述集船员位置管理及船员健康管理的可穿戴式终端设备配戴在船员身上,用于获取船员数据包括船员健康监测信息和实时位置信息,并通过蓝牙网关将数据信息回传数据动态分析模块;所述红外热成像人脸识别测温设备,运用非接触全幅人体测温方式对人体表面温度进行远距离非接触性测温,利用高清像素网络摄像机采集高清人脸图像,并配套使用动态人脸识别算法智能识别船员身份和姓名,实现动态、高速、多人脸检测、抓拍和识别;监测的数据传输至数据动态分析模块;所述数据动态分析模块,对船员数据和船舶数据进行统计分析;所述警示模块根据数据动态分析模块的数据分析,提供船舶和船员实时报警;所述人为审核端针对异常数据进行人为审核,当监测值超出预设阈值时发出警报至警示模块;通过管理人员人工核查或派人员上船稽查,进一步核实数据动态分析模块中监测异常的数据,保证监测数据的合理性和真实性;所述大型船舶安全监管系统终端包括船舶管理子系统和船员管理子系统,提供岸端管理者查看并管理船舶信息、船员信息;所述实时指导模块提供大型船舶远程作业的可视化分析和施工指导,并基于虚拟现实技术三维重现大型船舶状况;所述船舶信息数据库,用于存储大量船舶信息、船员不同状况的身体数据和实时位置数据,实现对船舶及其人员的历史数据回溯。
- 如权利要求2或者3所述的大型船舶安全监管系统,其特征在于,所述蓝牙网关,分为船舶室内和室外蓝牙网关,具体为:船舶室内的蓝牙网关采用嵌入式整机式设计,内嵌wifi模块、以太网模块以及蓝牙收发模块,能穿透大型船舱内钢板,满足数据传输需求;船舶室外设计的蓝牙网关具备防尘、防水、防雷、坚固耐用的特点,适用于特定大型船舶室外场景,支持POE供电;通过蓝牙网关组成扫描矩阵,并经POE供电或连接至交换机,将扫描到的蓝牙标签信号传输到数据动态分析模块。
- 如权利要求2所述的大型船舶安全监管系统,其特征在于,所述数据动态分析模块:船员数据包含船舶人员的身体健康数据和实时舱内位置数据,针对不同时段监测的船员数据进行动态统计分析;系统通过对船员的动态实时监测,保证船员的安全位置及身体健康;同时,数据动态分析模块定期自动生成船员身体健康状况数据报告并反馈船舶人员;船舶数据包括船舶航行/施工全过程的数据、特定船舶生产状态和设备运行状态的数据,以及船舶周边海况信息数据;船舶数据由传感器自动采集并采用无线传输方式实时传送至数据动态分析模块;数据动态分析模块将不同的国家和地区船舶的施工状况、设备运行状况、海事警报和海洋环境安全因素进行集 中监控分析。
- 如权利要求2所述的大型船舶安全监管系统,其特征在于,所述警示模块:船舶报警涉及远程船舶的实时安全监控、安全检查和安全管理,包含提供实时船舶预警警报和实时海事警告;船员报警包括主动报警和被动报警;主动报警指船员主动触发集船员位置管理及船员健康管理的可穿戴式终端设备的SOS按键,主动发出告警并在终端实时显示告警的标签、人员及实时舱内位置;被动报警指体温、睡眠、运动、血压、心率、血氧这类健康数据异常警报,以及船员进入/离开/滞留非访问区域或特定区域,误闯危险区域报警。
- 如权利要求2所述的大型船舶安全监管系统,其特征在于,所述大型船舶安全监管系统终端:其船舶管理子系统实现对船舶安全生产活动的记录、取证、报警安全管理功能;利用多维信息融合技术、HTML5及WebGL技术将施工现场船舶数据通过互联网发布,实现海洋环境工况、船舶状态、特殊作业生产状态、船舶设备位置及运行状态的全面监控;结合数据走势图、多倍速的历史回放功能,提供简便的数据信息共享服务,实现远程的施工监控、船机运行监控和故障诊断,及时调整作业方案,实现远程决策;其船员管理子系统通过给大型船舶作业人员配备可穿戴式终端设备来实现对作业人员的位置管理,实时监控管理船员室内位置、轨迹跟踪,具备一键报警、安全电子围栏、误闯危险区域报警功能;通过在后台管理中设置电子围栏和SOS一键紧急报警,对非访问区域或特定区域进行电子围栏设置,进入/离开/滞留时系统立即告警,保障人员及区域安全;系统同时实现对大型船舶作业人员的基本信息的管理维护、健康状况信息管理等,提供体温监测、睡眠监测、运动监测、血压监测服务,具备体温筛查及警报,辅助疫情期间的船员管理和疫情追溯。
- 如权利要求2所述的大型船舶安全监管系统,其特征在于,基于模型分 类的混合自适应压缩技术对数据进行压缩:首先建立多分类器进行分类,分类模型的输入层、输出层如下:
输入层={数据波动频率,数据类型}
输出层={0,1,2,3}其中,输入层的数据波动频率指采集到数据的波动频率,数据类型是指船舶数据、船员数据、水文数据、气象数据、海况数据;分类模型根据输入自动选择输出,具体表现为:S1:适用于船舶抛锚停工场景下,数据波动频率无波动时,输出结果倾向为0;S2:数据波动频率缓慢即数值在100秒及以上时间只有±3%以内波动、数据类型为船舶数据、水文数据时,输出结果倾向为1;S3:数据波动频率较慢即数值在30秒及以上时间无变化,且数据类型为气象数据、海况数据时,输出结果倾向为2;S4:针对除S1、S2或S3的其他大部分的情形,倾向输出结果为3;根据模型的输出结果{0,1,2,3}自动选择压缩方法,具体为:S1:当输出为0时,采用的压缩方法:动态设置时间窗口大小Wt,利用时间窗口分割数据,得到N组数据,保留前N-1组数据的第一个数据点,和第N组数据的最后一个数据点,实现高效压缩;S2:当输出为1时,采用的压缩方法:将一个相同值的连续串用其值和串长的数对二元组来替代,解压缩时根据字符及连续相同字符的个数可恢复至原来的数据;S3:当输出为2时,采用的压缩方法为以下四个步骤:步骤31:在采集的数据输入流中读取一个字符,进入步骤32;步骤32:如果当前编码在字典中,则把当前编码的第一个字符作为当前字符串的后缀,如果当前字符串不在字典中,就把它加入到字典中,然后把当前编码作为字符串的前缀,转到步骤34;步骤33:如果当前编码不在字典中,就把前缀的第一个字符作为后缀,把 字符串加入到字典中,用当前串的编码作前缀,转到步骤34;步骤34:把前缀放到输出流,转到步骤31;S4:当输出为3时,采用基于均方误差的旋转门压缩算法(Mean-based Spinning Door Transformation,简称MSDT);算法具体包含以下七个步骤:步骤41:(t0,y0)是上一个存储点,Δ是压缩偏差参数,kupper和klower分别是上一个存储点和新接收测试点之间斜率的上、下限;初始化a=0,b=0,c=0,kupper=+∞,klower=-∞;步骤42:当接收到新测点(t,y),计算步骤43:如果klower<k<kupper,转到步骤45继续,否则,执行下一步;步骤44:该点未通过压缩测试,被存储并被用作新的上一个存储点,然后重新初始化三个参数a,b,c:
a=τ2
b=-2kτ2
c=k2τ2-Δ2这里,τ和k的值利用新的上一个存储点进行计算,然后转到步骤46继续;步骤45:迭代计算a,b,c:
步骤46:如果b2-4ac≤0,该点未通过压缩测试,被存储并将其作为新的上一个存储点,然后转到步骤41继续压缩数据;步骤47:计算上限kupper和下限klower:
然后,转到步骤42继续压缩数据。 - 如权利要求1所述的大型船舶安全监管系统,其特征在于,船、岸两端数据传输独创基于模糊神经网络的数据传输链路智能优选技术,实现船、岸两端移动通信链路和卫星宽带链路的自动切换;基于模糊神经网络的数据传输链路智能优选技术根据当前网络参数计算并选择最优链路进行传输,以适应不同的施工环境,能在尽量降低成本的同时增加传输稳定性;基于模糊神经网络的数据传输链路智能优选技术的初始输入为{网络链路的接收信号强度RSS、带宽、负载、延时、信噪比、通信费用},最终输出为{移动通信,卫星宽带};基于模糊神经网络的数据传输链路智能优选技术包含六层设计,分别为输入层、模糊化层、模糊规则层、模糊决策层、解模糊层和输出层,具体为:输入层:输入层设置六个输入参数,所述六个输入参数为网络链路属性参数,分别为接收信号强度RSS、网络链路带宽、负载、延时、信噪比、通信费用;模糊化层:根据隶属度函数将从输入层接收到的网络链路属性参数模糊化,得到模糊集合,网络链路的接收信号强度RSS、带宽、负载、延时、信噪比和通信费用被映射成为{低,中,高}的模糊集合;模糊规则层:用来计算候选链路的优劣等级,利用神经网络将候选链路等级划分为{极好,较好,可以,不好,较差,极差}六大类;模糊决策层:利用神经网络分类器搭建六输入、六输出的神经网络架构;给定含N个初始训练样本的数据集{X,Y}={(xi,ym)|i,m=1,2,....,6,X∈R6,Y∈R6},设计神经网络的训练过程公式如下:
其中xi代表输入值,L为隐藏层神经元个数,wj代表输入层神经元连接第j个隐藏层神经元的权值,Bj代表第j个隐藏层神经元连接输出层神经元的权值,gj()代表激活函数,bj代表隐藏层偏置,N代表样本个数,yi代表输出值;神经网络分类的输入层向量和输出层向量分别为:输入层={RSS、带宽、负载、延时、信噪比、通信费用}输出层={极好,较好,可以,不好,较差,极差}解模糊层:模糊决策层将候选链路等级的自然语言变量定义为{极好,较好,可以,不好,较差,极差},解模糊层将自然语言模糊集合转化为精确数值用于后续的比较;解模糊化的计算公式如下:
其中p取值在[1,2],μ(x)表示隶属度函数,c和σ分别为隶属函数的中心和标准差,Sl表示链路最终得分;输出层:在获得的各个链路的得分值之后进行比较,选取Sl中的最高值,并自动切换到得分最高的链路进行数据传输。基于模糊神经网络的数据传输链路智能优选技术选取具有最大链路得分值的候选链路,在减少计算开销的情况下提高链路选择结果的准确性,使得系统能够始终实时地自动根据当前网络参数计算并选择最优链路进行传输,以适应不同的施工环境。 - 如权利要求2所述的大型船舶安全监管系统,其特征在于,所述船舶信息数据库:系统针对每艘船舶和每位船员分别建立独立的档案数据;在大型船舶有联 网情况下,由船端服务器向岸端服务器推送数据,从而同步船上服务器的数据,并进行数据查看;通过配置HTTP推送路径方式,由本地服务器同步推送至云端服务器,实现船、岸、云端信息同步。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2023/132323 WO2024104461A1 (zh) | 2023-11-17 | 2023-11-17 | 一种大型船舶安全监管系统 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2023/132323 WO2024104461A1 (zh) | 2023-11-17 | 2023-11-17 | 一种大型船舶安全监管系统 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2024104461A1 true WO2024104461A1 (zh) | 2024-05-23 |
Family
ID=91083869
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2023/132323 WO2024104461A1 (zh) | 2023-11-17 | 2023-11-17 | 一种大型船舶安全监管系统 |
Country Status (1)
Country | Link |
---|---|
WO (1) | WO2024104461A1 (zh) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114777034A (zh) * | 2022-04-29 | 2022-07-22 | 东方通用航空摄影有限公司 | 一种天然气长输管道应急管理云平台及方法 |
CN118298514A (zh) * | 2024-06-06 | 2024-07-05 | 华东交通大学 | 一种基于深度学习的工人危险动作识别方法及系统 |
CN118410344A (zh) * | 2024-07-04 | 2024-07-30 | 中国船舶集团有限公司第七〇七研究所 | 一种基于模糊神经网络的船舶航行安全效能评估方法 |
CN118524444A (zh) * | 2024-07-24 | 2024-08-20 | 中交上海航道装备工业有限公司 | 一种无线通信多链路负载均衡与智能优化方法 |
CN118574078A (zh) * | 2024-07-31 | 2024-08-30 | 杭州北岑科技有限公司 | 基于防管控应用的监测方法、设备、介质及程序产品 |
CN118691050A (zh) * | 2024-08-26 | 2024-09-24 | 武汉中科坤程信息产业有限公司 | 一种用于智慧文旅的共享船舶控制系统 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112255946A (zh) * | 2020-10-19 | 2021-01-22 | 青岛博瑞斯自动化技术有限公司 | 一种基于云服务和大数据的船舶状态远程监控系统及方法 |
CN113408941A (zh) * | 2021-07-12 | 2021-09-17 | 大连海事大学 | 一种服务于智能船舶的大数据集成平台 |
KR20210138985A (ko) * | 2020-05-13 | 2021-11-22 | 대우조선해양 주식회사 | 선원 건강 관리 시스템 및 방법 |
KR20230040019A (ko) * | 2021-09-15 | 2023-03-22 | 대우조선해양 주식회사 | 선박 알람정보 모니터링 시스템 및 동 시스템을 구비한 선박 |
KR102524953B1 (ko) * | 2022-11-07 | 2023-04-25 | 한국해양과학기술원 | 선내시스템 원격모니터링 서비스 장치 및 방법 |
-
2023
- 2023-11-17 WO PCT/CN2023/132323 patent/WO2024104461A1/zh unknown
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20210138985A (ko) * | 2020-05-13 | 2021-11-22 | 대우조선해양 주식회사 | 선원 건강 관리 시스템 및 방법 |
CN112255946A (zh) * | 2020-10-19 | 2021-01-22 | 青岛博瑞斯自动化技术有限公司 | 一种基于云服务和大数据的船舶状态远程监控系统及方法 |
CN113408941A (zh) * | 2021-07-12 | 2021-09-17 | 大连海事大学 | 一种服务于智能船舶的大数据集成平台 |
KR20230040019A (ko) * | 2021-09-15 | 2023-03-22 | 대우조선해양 주식회사 | 선박 알람정보 모니터링 시스템 및 동 시스템을 구비한 선박 |
KR102524953B1 (ko) * | 2022-11-07 | 2023-04-25 | 한국해양과학기술원 | 선내시스템 원격모니터링 서비스 장치 및 방법 |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114777034A (zh) * | 2022-04-29 | 2022-07-22 | 东方通用航空摄影有限公司 | 一种天然气长输管道应急管理云平台及方法 |
CN118298514A (zh) * | 2024-06-06 | 2024-07-05 | 华东交通大学 | 一种基于深度学习的工人危险动作识别方法及系统 |
CN118410344A (zh) * | 2024-07-04 | 2024-07-30 | 中国船舶集团有限公司第七〇七研究所 | 一种基于模糊神经网络的船舶航行安全效能评估方法 |
CN118410344B (zh) * | 2024-07-04 | 2024-10-15 | 中国船舶集团有限公司第七〇七研究所 | 一种基于模糊神经网络的船舶航行安全效能评估方法 |
CN118524444A (zh) * | 2024-07-24 | 2024-08-20 | 中交上海航道装备工业有限公司 | 一种无线通信多链路负载均衡与智能优化方法 |
CN118574078A (zh) * | 2024-07-31 | 2024-08-30 | 杭州北岑科技有限公司 | 基于防管控应用的监测方法、设备、介质及程序产品 |
CN118574078B (zh) * | 2024-07-31 | 2024-10-18 | 杭州北岑科技有限公司 | 基于防管控应用的监测方法、设备、介质及程序产品 |
CN118691050A (zh) * | 2024-08-26 | 2024-09-24 | 武汉中科坤程信息产业有限公司 | 一种用于智慧文旅的共享船舶控制系统 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2024104461A1 (zh) | 一种大型船舶安全监管系统 | |
CN109447048B (zh) | 一种人工智能预警系统 | |
CN105913559B (zh) | 一种基于体感技术的银行atm机智能监控方法 | |
CN108846332B (zh) | 一种基于clsta的铁路司机行为识别方法 | |
US7242295B1 (en) | Security data management system | |
CN110133573A (zh) | 一种基于多元传感器信息融合的自主低空无人机防御系统 | |
CN110996067A (zh) | 基于深度学习的高危作业环境下的人员安全实时智能视频监控系统 | |
CN111383132A (zh) | 一种适用于疫情期的智慧工地管理系统及管理方法 | |
CN116563797B (zh) | 一种用于智慧校园的监控管理系统 | |
CN110135564A (zh) | 基于边缘计算的煤矿瓦斯传感器数据异常检测方法 | |
CN117493969A (zh) | 一种大型船舶安全监管系统 | |
CN115761537B (zh) | 一种面向动态特征补充机制的输电线路异物入侵识别方法 | |
CN110525456A (zh) | 一种列车安全驾驶监测系统及方法 | |
CN115457446A (zh) | 一种基于视频分析的异常行为监管系统 | |
CN117319609A (zh) | 一种物联网大数据智能视频监控系统及方法 | |
CN113642403B (zh) | 基于边缘计算的人群异动智能安全检测系统 | |
CN108256447A (zh) | 一种基于深度神经网络的无人机航拍视频分析方法 | |
CN111553264B (zh) | 一种适用于中小学生的校园非安全行为检测及预警方法 | |
CN211184122U (zh) | 铁路作业安全防控和大客流预警联动的智能视频分析系统 | |
CN104349129A (zh) | 基于数字化信息技术的校园智能安全监控系统 | |
CN118433231A (zh) | 基于边缘移动监测站的智慧工地施工安全监测云边协同预警系统及其控制系统及其控制方法 | |
CN113705364B (zh) | 基于人工智能的输电线路外部隐患预警系统及方法 | |
CN113159984B (zh) | 一种变电站工作人员工作路径追踪方法 | |
CN116246424B (zh) | 一种老年人行为安全监控系统 | |
CN117912190A (zh) | 一种防欺凌智能手表系统 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 23890890 Country of ref document: EP Kind code of ref document: A1 |