Ship lock control system process safety detection and early warning method

Abstract

The invention discloses a ship lock control system process safety detection and early warning method, and relates to the technical field of ship lock process safety, and the method comprises the steps: S1, data collection unit construction and data collection analysis; the data acquisition unit can comprehensively acquire key process data of the ship lock, the CPU module accurately analyzes and judges potential safety hazards and outputs early warning, the digital quantity input module achieves wide data acquisition, the output module provides protection signals, and the communication module guarantees data communication of all the units. The storage and forwarding unit realizes data decoding, storage and efficient forwarding through a code scanning gun and a two-dimensional code display screen, an intelligent gateway ensures stable transmission of data, an interactive touch screen is convenient for an operator to check a system state, set parameters and query data, and the whole process safety system is independent of an industrial control system, so that the operation safety and reliability of the ship lock are enhanced; the defects of an existing ship lock control system in the aspects of data acquisition and analysis, communication storage, interaction functions, safety guarantee and the like are effectively overcome.

Description

Technical Field

[0001] The present invention relates to the technical field of ship lock process safety, and in particular to a method for process safety detection and early warning of a ship lock control system. Background Art

[0002] In modern water transportation, locks play a vital role. The lock control system has undergone a transformation from manual operation to automated control. Initially, the operation of locks mainly relied on manual control, which was not only inefficient but also posed a great safety hazard. With the continuous development of automation technology, lock control systems have gradually realized automated control. Through the integrated control system, functions such as lock gate control, water level monitoring, and electromechanical control are realized, thereby improving the efficiency and accuracy of lock operation and reducing the labor intensity of operators. The lock process is complex, involving a variety of operation protection processes when ships pass through the navigation structure (lock) area, such as closing dislocation protection, water transfer time protection, dynamic water valve closing, intermittent valve opening, unilateral water transfer, etc. The safety of the lock process is directly related to the smooth passage of ships, the integrity of lock facilities, and the safety of life and property of personnel. By strengthening safety management and technical application in the design, construction, commissioning, operation and maintenance links, the safe operation and efficient scheduling of locks can be ensured, providing strong support for the development of the water transportation system.

[0003] However, traditional ship lock control systems have many shortcomings in terms of process safety detection and early warning. Existing data collection methods are often relatively simple and inaccurate, unable to fully and real-timely obtain various data on the key control processes of the ship lock, making it difficult to accurately determine whether the lock's operating status contains safety hazards. The data transmission process is easily interfered with, and the transmission efficiency and reliability are low, resulting in the inability to timely and accurately transmit important safety information to the management unit. The lack of effective data storage management strategies makes it difficult to quickly retrieve and utilize historical data for safety analysis. The incomplete interactive function makes it difficult for operators to conveniently obtain system operating status, adjust parameters, and query data. These problems seriously affect the safety and management efficiency of ship lock operations. Therefore, it is necessary to propose a method for process safety detection and early warning of ship lock control systems to address the problems in the existing technology. Summary of the Invention

[0004] The purpose of the present invention is to make up for the shortcomings of the existing technology and provide a method for process safety detection and early warning of the lock control system. It can be independent of the industrial control system, enhance the safety and reliability of the lock operation, and effectively solve the shortcomings of the existing lock control system in data acquisition and analysis, communication storage, interactive functions and security assurance.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for process safety detection and early warning of a ship lock control system, the method comprising the following steps:

[0006] S1. Data acquisition unit construction and data collection and analysis:

[0007] The data acquisition unit consists of a CPU module, a digital input module, a digital output module, and a data communication module. The digital input module collects various physical quantity data during the key control process of the ship lock, including but not limited to valve opening data, water level data, and equipment operating status data. The collected data is encoded in a specific format and transmitted to the CPU module.

[0008] The CPU module processes data using a self-developed security assessment algorithm based on dynamic thresholds and trend analysis. The algorithm formula is: Where S is the safety assessment value, x i is the data related to the i-th valve, y j is the j-th water level related data, w i and v j is the corresponding weight coefficient, f(x i ) and g(y j ) is a function defined based on the dynamic change characteristics of the data. Its specific form is determined according to historical data and the operating characteristics of the ship lock. By calculating the value of S and comparing it with the preset safety threshold range, it is determined whether there are safety hazards in the key control process of the ship lock. If it exceeds the threshold range, an early warning information is generated. The early warning information contains detailed information about the type of safety hazard, the possible affected area, and the recommended measures to be taken. At the same time, the data is marked as abnormal data to be detected and enters the detection process;

[0009] The data communication module adopts the CAN bus communication protocol based on time division multiplexing and error correction coding to realize data communication between various data acquisition units;

[0010] S2. Data processing and transmission of data storage and forwarding unit:

[0011] The data acquisition unit encodes and outputs the collected data in the form of a QR code through the 485 protocol screen, and the barcode scanner in the storage and forwarding unit scans the QR code to obtain the data;

[0012] The intelligent gateway in the storage and forwarding unit decodes and stores the acquired data. The storage format adopts a hybrid storage structure based on data blocks and timestamps. The intelligent gateway uses its own adaptive data compression and transmission algorithm to connect the data to the communication switch via Ethernet. The algorithm dynamically adjusts the compression ratio according to the network load and data importance. The formula is: Where C is the compression ratio, z iis the i-th important data item, t j is the jth non-important data item, k is the adjustment coefficient, N is the current network load, and T is the preset time period; the intelligent gateway and the process safety communication switch in the integrated cabinet of the sub-control center form a ring network structure to realize the transmission of field-collected data to the upper-level management unit;

[0013] S3. Functional implementation of interactive touch screen:

[0014] The interactive touch screen is connected to the data acquisition unit and the storage and forwarding unit to obtain system working status information in real time;

[0015] The interactive touch screen provides a parameter setting interface for the data acquisition unit. Operators can enter the setting interface through specific touch gestures and password verification mechanisms to adjust the data acquisition frequency and warning threshold parameters. The parameter adjustment instructions are transmitted to the data acquisition unit through a custom secure communication protocol.

[0016] The interactive touch screen has a data query function. It can retrieve and display the corresponding historical data from the data stored in the intelligent gateway according to the time range and data type query conditions entered by the operator. The display format includes charts and lists. The chart type automatically adapts to the data characteristics, including line charts to show the change trend of water level over time and bar charts to show the statistics of valve opening and closing times.

[0017] S4. Abnormal data detection and early warning:

[0018] For data marked as abnormal, a detection method based on multivariate statistical analysis and machine learning models was used. Multiple relevant data variables, including valve actuation speed, water level change rate, and equipment vibration frequency, were selected to construct a data matrix. Principal component analysis (PCA) was then used to reduce the dimensionality of the data matrix and extract key features.

[0019] The extracted features are input into pre-trained support vector machines and neural network machine learning models for classification and judgment to determine whether the data is truly abnormal and the severity of the abnormality. If it is determined to be abnormal and the severity meets the warning standard, an alert will be issued through various means, including popping up a prominent warning box on the interactive touch screen to display detailed information about the abnormality;

[0020] At the same time, SMS and email notifications are sent to relevant management personnel. The warning information includes detailed analysis results of abnormal data, possible security risks and response suggestions. In addition, the collection parameters of the data collection unit are automatically adjusted according to the abnormal situation, including increasing the collection frequency and expanding the collection range.

[0021] Furthermore, in the CPU module of the data acquisition unit, the weight coefficient w in the security assessment algorithm based on dynamic threshold and trend analysis isi and v j The method for determining is as follows:

[0022] Regularly collect historical data of the lock under normal operation and simulated fault conditions, and pre-process the data, including removing outliers and normalizing the data;

[0023] According to the valve opening data, the weight coefficient w is determined by combining the analytic hierarchy process (AHP) and the fuzzy comprehensive evaluation method according to the degree of influence of different opening ranges on the safe operation of the ship lock. i First, a judgment matrix is ​​constructed. The matrix elements are determined by expert experience and data analysis to reflect the relative importance of different opening data. Then, the judgment matrix is ​​fuzzified using the fuzzy comprehensive evaluation method to obtain a fuzzy relationship matrix. Finally, the weight vector w is obtained through fuzzy synthesis operation. i ;

[0024] For the water level height data, according to the design water level standard of the ship lock and the actual operating water level fluctuation range, the weight coefficient is determined by combining the variation coefficient method and the entropy weight method, and the variation coefficient v of each water level data is calculated. j , reflecting the degree of data dispersion, and calculating the information entropy at the same time. The weight contribution of the data is determined according to the information entropy, and the weight coefficient v is obtained by combining the two. j .

[0025] Furthermore, the digital input module in the data acquisition unit uses multi-sensor redundant acquisition technology when collecting data, specifically:

[0026] Arrange multiple sensors of the same type at key locations, including multiple opening sensors at valves and multiple water level sensors at water level measurement points;

[0027] Each sensor collects data independently and transmits the data to the data acquisition unit;

[0028] The CPU module in the data acquisition unit uses a self-created algorithm based on data consistency check and weighted fusion to process the data collected by multiple sensors. The formula is: Where D is the fused data, d k is the data collected by the kth normal sensor, e l The data collected by the lth suspected fault sensor, a k and b l is the corresponding weighting coefficient, which is dynamically determined based on the historical reliability data of the sensor and the current data consistency. If the difference in data collected by multiple sensors exceeds the preset consistency threshold, it is judged that the sensor may be faulty, and its weighting coefficient is reduced or its data is directly excluded. Normal sensor data is used for fusion calculation.

[0029] Furthermore, in the CAN bus communication protocol based on time division multiplexing and error correction coding adopted by the data communication module, the time slot allocation method of time division multiplexing is as follows:

[0030] Divide the communication time slots into different time slot groups according to the data type and transmission priority collected by the data acquisition unit;

[0031] For critical data with high real-time requirements, including emergency valve closing instructions and fault alarm signals, more time slot resources are allocated for timely transmission;

[0032] For common data, including equipment operation status monitoring data, relatively few time slot resources are allocated;

[0033] The time slot allocation is dynamically adjusted according to the operating status of the lock. When the lock is in frequent operation or emergency, the time slot requirements of critical data are prioritized. This dynamic time slot allocation method improves the communication efficiency of the CAN bus and the real-time performance of data transmission.

[0034] Furthermore, in the intelligent gateway of the storage and forwarding unit, the method for determining the adjustment coefficient in the adaptive data compression and transmission algorithm is as follows:

[0035] Real-time monitoring of network performance indicators such as network bandwidth utilization, data transmission delay, and data loss rate;

[0036] Based on these indicators, a comprehensive evaluation function is constructed. The function form is: E = α × U + β × D + γ × L, where E is the comprehensive evaluation value of network performance, U is the network bandwidth utilization, D is the data transmission delay, L is the data loss rate, and α, β, and γ are weight coefficients determined according to the network performance requirements of the ship lock control system.

[0037] According to the range of the comprehensive evaluation value E, the value of the adjustment coefficient k is determined in the form of a piecewise function. When E is small, it indicates that the network performance is good. Increasing the k value appropriately improves the data compression ratio and reduces the data transmission volume. When E is large, it indicates that the network performance is poor. Reducing the k value reduces the compression ratio. By dynamically adjusting the adjustment coefficient k in this way, the intelligent gateway can optimize the data compression and transmission strategy according to the actual network situation.

[0038] Furthermore, the touch operation gesture recognition of the interactive touch screen adopts a gesture recognition model based on deep learning. The specific steps are as follows:

[0039] Collect touch gesture samples from operators under different environmental conditions, including click, slide, and zoom gestures, and annotate the gesture samples with gesture type, start position, end position, and duration information;

[0040] Construct a convolutional neural network (CNN) model. The model structure includes convolutional layers, pooling layers, and fully connected layers. The convolutional layers are used to extract features of gesture images, the pooling layers are used to reduce data dimensions, and the fully connected layers are used to classify gesture types.

[0041] The CNN model is trained using the labeled gesture samples. During the training process, the stochastic gradient descent algorithm is used to optimize the model parameters. By continuously adjusting the parameters, the model can accurately recognize different touch operation gestures.

[0042] The trained model is deployed on the interactive touch screen. When the operator performs a touch operation, the touch screen collects gesture images, which are pre-processed and input into the model for recognition. Based on the recognition results, corresponding operations are performed, including opening the parameter setting interface and querying historical data.

[0043] Furthermore, the secure communication protocol between the interactive touch screen and the data acquisition unit and the storage and forwarding unit adopts an encryption and authentication mechanism based on elliptic curve encryption ECC and message authentication code MAC, specifically:

[0044] Before data transmission, the sender uses the elliptic curve encryption algorithm to encrypt the transmitted data and generate ciphertext;

[0045] At the same time, the sender uses the message authentication code algorithm to calculate the MAC value of the data and transmits the MAC value together with the ciphertext;

[0046] After receiving the data, the receiver first uses the same elliptic curve encryption algorithm to decrypt the ciphertext to obtain the original data;

[0047] The receiver then uses the same message authentication code algorithm to calculate the MAC value of the received data and compares it with the MAC value sent by the sender. If the two are consistent, the data is considered to have not been tampered with and the source is reliable, and the data is accepted. Otherwise, the data is rejected and an alarm is issued.

[0048] Furthermore, when outputting protection signals, the digital output module in the data acquisition unit adopts a signal output strategy based on priority and delay time, specifically:

[0049] Different priorities are set for different types of protection signals, including shutdown protection signal with the highest priority, emergency valve closing protection signal with the second highest priority, and other protection signals with lower priority;

[0050] When multiple protection signals are generated at the same time, they are output in order of priority;

[0051] For each protection signal, a corresponding delay time is set according to the response characteristics and safety requirements of the lock equipment. If a higher priority protection signal is generated during the delay time, the output of the current protection signal will be suspended and the higher priority protection signal will be output first.

[0052] The delay time is set based on the equipment dynamics model and safety time window analysis. By establishing a dynamic model of the lock equipment, analyzing the response time of the equipment under different working conditions, and combining the safety time window requirements to determine a reasonable delay time.

[0053] Furthermore, in the intelligent gateway of the storage and forwarding unit, the hybrid storage structure based on data blocks and timestamps for data storage is specifically as follows:

[0054] The collected data is divided into different data blocks according to data type and collection time. Each data block contains the same type of data within a certain time range.

[0055] Add timestamp information to the header of each data block to record the start and end time of the data block;

[0056] For frequently accessed data blocks, including recent key operating data, high-speed cache technology is used to store them to improve data query speed;

[0057] For historical data blocks, hierarchical storage is performed based on the importance and access frequency of the data. Highly important and frequently accessed data blocks are stored in high-performance storage media, while other data blocks are stored in high-capacity, low-cost storage media.

[0058] Create a data index table that records the storage location, data type, and time range information of the data block to facilitate data query and management.

[0059] Compared with the existing technology, the method of process safety detection and early warning of the lock control system has the following beneficial effects:

[0060] The data acquisition unit of the present invention can comprehensively collect key process data of the lock, the CPU module accurately analyzes and determines safety hazards and outputs early warnings, the digital input module realizes extensive data acquisition, the output module provides protection signals, the communication module ensures data communication between each unit, and the storage and forwarding unit realizes data decoding, storage and efficient forwarding through a barcode scanner and a QR code display screen. The intelligent gateway ensures stable data transmission, and the interactive touch screen facilitates operators to view system status, set parameters and query data. The overall process safety system is independent of the industrial control system, which enhances the safety and reliability of lock operation and effectively solves the shortcomings of existing lock control systems in data acquisition and analysis, communication storage, interactive functions and safety assurance.

[0061] Other advantages, objects and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be learned from the practice of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0063] Figure 1 A schematic diagram of the framework structure of a method for process safety detection and early warning of a ship lock control system according to the present invention;

[0064] Figure 2 Schematic diagram of the framework structure of the data acquisition unit in the present invention;

[0065] Figure 3 Schematic diagram of the framework structure of the storage and forwarding unit in the present invention;

[0066] In the accompanying drawings, the components represented by the reference numerals are as follows:

[0067] 1-Data acquisition unit, 11-CPU module, 12-Digital input module, 13-Digital output module, 14-Data communication module, 2-Storage and forwarding unit, 21-Barcode scanner, 22-QR code display, 23-Intelligent gateway, 3-Communication switch, 4-Interactive touch screen. DETAILED DESCRIPTION

[0068] The following is a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0069] Example 1

[0070] A large ship lock is located on a busy waterway with many ships passing through every day. The lock is equipped with an advanced control system and adopts the process safety detection and early warning method of the present invention. Before the lock is operated, the operator enters the parameter setting interface of the acquisition unit through the interactive touch screen and uses specific touch operation gestures and password verification mechanism (based on the gesture recognition model of deep learning to ensure the safety and convenience of operation). According to the design specifications of the lock and past operating experience, the data collection frequency is set to collect key data such as valve opening and water level every 0.5 seconds, and a reasonable early warning threshold is set. For example, an early warning is triggered when the water level deviates from the normal range by ±0.5 meters. At the same time, according to the characteristics of the lock equipment, the priority of different protection signals of the digital output module in the data acquisition unit is determined, such as the shutdown protection signal has the highest priority, the emergency valve closing protection signal has the second highest priority, and the corresponding delay time (set by analysis based on the equipment dynamics model and the safety time window).

[0071] The digital input module in the data acquisition unit (using multi-sensor redundant acquisition technology, arranging multiple sensors of the same type at key locations such as valves and water level measurement points) begins to collect data from the key control process of the lock. For example, multiple opening sensors at the valve and multiple water level sensors at the water level measurement point work simultaneously, independently collecting data and transmitting it to the data acquisition unit.

[0072] The CPU module receives the collected data and processes it using a safety assessment algorithm based on dynamic thresholds and trend analysis (the weight coefficient is determined by a complex calculation method, such as a combination of hierarchical analysis and fuzzy comprehensive evaluation for valve opening data, and a combination of coefficient of variation and entropy weight for water level data). Assuming that the currently collected valve opening data are x1, x2, ..., x n , the water level data is y1,y2,…,y m , according to the algorithm formula Calculate the safety assessment value S, where f(x i ) and g(y j ) is a function determined based on historical data and lock operation characteristics, and is used to highlight the dynamic change characteristics of the data.

[0073] If the calculated safety assessment value S exceeds the preset safety threshold range, the CPU module immediately generates an early warning message, including details such as the type of safety hazard (such as valve abnormality or water level abnormality), possible affected areas (such as affecting the normal operation of a certain lock chamber), and recommended measures (such as checking the valve drive device or checking the upstream and downstream water level adjustment). At the same time, the relevant data is marked as abnormal data to be detected, ready to enter the next step of the detection process.

[0074] The data acquisition unit encodes and outputs the collected data in the form of a QR code through the 485 protocol screen, and the barcode scanner in the storage and forwarding unit scans the QR code to obtain the data.

[0075] The intelligent gateway decodes and stores the acquired data, using a hybrid storage structure based on data blocks and timestamps (frequently accessed data blocks such as recent critical operating data are stored in the cache, historical data is stored hierarchically based on importance and access frequency, and a data index table is established).

[0076] The intelligent gateway uses an adaptive data compression and transmission algorithm (the adjustment coefficient k is dynamically determined based on real-time monitored network performance indicators such as network bandwidth utilization, data transmission delay and data loss rate) to connect data to the switch through Ethernet. The data communication module uses a CAN bus communication protocol based on time division multiplexing and error correction coding (time slots are dynamically allocated based on data type and transmission priority). The intelligent gateway and the process safety switch of the integrated cabinet in the sub-control center form a ring network structure, which transmits the data collected on-site to the upper-level management unit in a stable and efficient manner.

[0077] For data marked as abnormal data to be detected, a detection method based on multivariate statistical analysis and machine learning models is adopted. Multiple relevant data variables such as valve operation speed, water level change rate, and equipment vibration frequency are selected to construct a data matrix. Multivariate statistical analysis methods such as principal component analysis (PCA) are used to reduce the dimension and extract key features. The features are then input into a pre-trained support vector machine (SVM) model for classification and judgment. For example, it is determined whether the current anomaly is a short-term fluctuation or a serious anomaly caused by equipment failure.

[0078] If it is determined to be abnormal and the severity reaches the warning standard, an early warning will be issued in various ways. A prominent warning box will pop up on the interactive touch screen, displaying detailed information about the abnormality. At the same time, a text message notification will be sent to the mobile phone of the lock manager. The early warning information includes detailed analysis results of the abnormal data (such as abnormal trends discovered through data mining), possible safety risks (such as damage to lock equipment or risk of ship collision) and response suggestions (such as emergency stop of lock operation and inspection of related equipment). In addition, the collection parameters of the data acquisition unit will be automatically adjusted according to the abnormal situation, such as increasing the collection frequency to once every 0.2 seconds and expanding the collection range (such as increasing the collection of surrounding environmental data) to more accurately monitor the development trend of the abnormal situation.

[0079] Through the application of this embodiment, the lock achieves all-round and precise detection of process safety and timely and effective early warning during operation. The accuracy and comprehensiveness of data collection are greatly improved, and misjudgment caused by data errors is reduced. The efficient data storage and forwarding mechanism ensures the real-time transmission of information. The management unit can grasp the lock status and make decisions in a timely manner. The multi-mode early warning measures enable management personnel to respond quickly to abnormal situations and take effective measures to avoid accidents, thereby ensuring the safety of lock facilities and passing ships. At the same time, the convenient functions of the interactive touch screen and the independence of the system improve the efficiency and reliability of lock operation management, reduce maintenance costs, and improve the operation management level of the entire lock, providing a strong guarantee for the safety and smooth flow of busy waterways.

[0080] Example 2

[0081] A small ship lock is located on a tributary of an inland river, mainly serving the navigation needs of small cargo ships and cruise ships in the surrounding area. The lock introduced the control system process safety detection and early warning method of the present invention. In the early stage of the lock being put into use, technicians set the parameters of the data acquisition unit through the interactive touch screen. Since the small ship lock operates relatively infrequently, the data acquisition frequency is set to once every 2 seconds. At the same time, combined with the design water level limit and equipment performance of the lock, a more conservative early warning threshold is set. For example, an early warning is triggered when the water level height fluctuation exceeds ±0.3 meters or the valve action time exceeds the normal range of ±1 second. In the digital output module of the data acquisition unit, it is clarified that the emergency valve closing protection signal has the highest priority when the water level rises rapidly and exceeds the warning water level. The shutdown protection signal responds immediately when the equipment fails seriously. In addition, based on the relatively simple response characteristics of the lock equipment, a shorter delay time is set (determined by a simplified analysis of the equipment dynamics model and a conservative estimate of the safety time window).

[0082] The digital input module of the data acquisition unit begins to collect data. Although the ship lock is small in scale, multiple opening sensors are still arranged at key locations such as valves. Redundant water level sensors are also set up at water level measurement points to ensure data accuracy. The data collected by these sensors, such as the change in valve opening from 0 to 100% and the value of water level height relative to the reference surface, are transmitted to the data acquisition unit in real time.

[0083] After receiving the data, the CPU module processes it using a safety assessment algorithm based on dynamic thresholds and trend analysis. For small ship locks, considering their operating characteristics, when determining the weight coefficient, more emphasis is placed on the impact of water level data on safety assessment (through analysis of historical operating data of small ship locks, it is found that water level changes have a greater impact on ship lock safety). Assume that the collected water level data are y1, y2, ..., y m , the valve opening data is x1, x2,…, x n, according to the algorithm formula Calculate the safety assessment value S, where f(x i ) and g(y j ) function was adjusted specifically according to the operating characteristics of the small ship lock.

[0084] During a ship's passage through a lock, the water level began to rise rapidly due to a sudden rainstorm upstream. The rate of change of the water level data collected by the data acquisition unit exceeded the normal range. After calculation, the CPU module found that the safety assessment value S was close to the warning threshold. At this time, the CPU module closely monitored the data trend. When the water level continued to rise and the S value exceeded the warning threshold, an early warning message was immediately generated. The early warning message indicated that there might be a risk of flood impacting the lock, and it was recommended to make flood prevention preparations in advance, such as checking drainage facilities and strengthening the lock structure. At the same time, the relevant water level data was marked as abnormal data to be detected.

[0085] The data acquisition unit encodes the data into a QR code and outputs it through the 485 protocol screen. The barcode scanner of the storage and forwarding unit scans the QR code to obtain the data. This process ensures the safe and convenient transmission of data between the acquisition unit and the storage and forwarding unit, avoiding problems such as poor contact that may be caused by traditional cable connections.

[0086] The intelligent gateway decodes and stores the acquired data. Since the data volume of small ship locks is relatively small, the intelligent gateway adopts a relatively simple but efficient hybrid storage structure based on data blocks and timestamps, and stores recent water level and valve data in a high-speed cache to facilitate quick query and analysis. At the same time, according to the network environment of the small ship lock (relatively narrow but relatively stable bandwidth), the intelligent gateway uses an adaptive data compression and transmission algorithm (the adjustment coefficient is adjusted according to the network bandwidth utilization and data importance), accesses the switch through Ethernet, and the data communication module adopts the CAN bus communication protocol based on time division multiplexing and error correction coding to reasonably allocate time slots, give priority to the transmission of key data such as water level, and ensure that the data is stably transmitted to the process safety switch of the integrated cabinet in the sub-control center, forming a ring network structure to ensure the reliability of data transmission.

[0087] For the marked water level anomaly data, a detection method based on multivariate statistical analysis and machine learning models is adopted. Data variables such as the water level change rate, water level change amplitude, and correlation with water level changes in past rainstorms are selected to construct a data matrix. Principal component analysis (PCA) is used for dimensionality reduction. After extracting key features, they are input into a pre-trained neural network model for classification and judgment. After learning the historical water level anomaly data of the small ship lock, the neural network model can accurately determine whether the anomaly is caused by a short-term rainstorm or there may be other potential problems, such as abnormal water level rise caused by river blockage.

[0088] After testing, it was determined that the water level anomaly was caused by heavy rain and a small amount of debris accumulation in the river channel, and the severity reached the early warning standard. Therefore, a prominent warning box immediately popped up on the interactive touch screen, displaying detailed water level anomaly information, including the current water level, the rate of water level rise, and the predicted maximum water level. At the same time, an email notification was sent to the lock management personnel, which contained response suggestions based on data analysis, such as timely cleaning of river channel debris, adjustment of lock drainage strategies, etc. In addition, according to the abnormal situation, the collection parameters of the data acquisition unit were automatically adjusted, the water level data collection frequency was increased to once per second, and the collection of relevant data such as river flow velocity was increased to better monitor the water level change trend. Based on the early warning information, the management personnel quickly organized personnel to clean up the river channel debris and closely monitored the water level changes, ensuring the safe operation of the lock during the heavy rain.

[0089] In this embodiment, the process safety detection and early warning method of the present invention shows good adaptability and effectiveness. Through precise data collection and intelligent analysis, abnormal water levels are discovered in time and early warnings are issued, which provides valuable preparation time for responding to sudden rainstorms and floods. The efficient data storage, forwarding and transmission mechanism ensures the timely transmission of information. Although the lock is small in scale, the system still operates stably. The convenient operation function of the interactive touch screen makes it convenient for managers to obtain information and adjust parameters. The independence of the process safety system also ensures that the safety detection and early warning functions are not affected in the relatively simple industrial control system environment of small locks. On the whole, the ability of small locks to respond to emergencies such as natural disasters is improved, the safe operation of the locks is guaranteed, and the economic losses and navigation interruptions caused by safety accidents are reduced. At the same time, it also provides strong support for the intelligent management of small locks and improves their operational management level.

[0090] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and scope of the same elements of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

Claims

1. A method for process safety detection and early warning of a ship lock control system, characterized in that: The method comprises the following steps: S1. Data collection unit construction and data collection and analysis: The data acquisition unit is composed of a CPU module, a digital input module, a digital output module, and a data communication module. The digital input module collects various physical quantity data in the key control process of the ship lock, including but not limited to valve opening data, water level data, and equipment operation status data. The collected data is encoded in a specific format and transmitted to the CPU module; The CPU module processes the data using a self-created security assessment algorithm based on dynamic thresholds and trend analysis. The algorithm formula is: Where S is the safety assessment value, x i is the data related to the i-th valve, y j is the j-th water level related data, w i and v j is the corresponding weight coefficient, f(x i ) and g(y j ) is a function defined based on the dynamic change characteristics of data. Its specific form is determined according to historical data and the operation characteristics of the ship lock. By calculating the value of S and comparing it with the preset safety threshold range, it is determined whether there are safety hazards in the key control process of the ship lock. If it exceeds the threshold range, an early warning information is generated. The early warning information includes the type of safety hazard, the possible affected area and the detailed content of the recommended measures. At the same time, the data is marked as abnormal data to be detected and enters the detection process; The data communication module adopts the CAN bus communication protocol based on time division multiplexing and error correction coding to realize data communication between various data acquisition units; S2. Data processing and transmission of data storage forwarding unit: The data acquisition unit encodes and outputs the collected data in the form of a QR code through the 485 protocol screen, and the barcode scanner in the storage and forwarding unit scans the QR code to obtain the data; The intelligent gateway in the storage and forwarding unit decodes and stores the acquired data. The storage format adopts a hybrid storage structure based on data blocks and timestamps. The intelligent gateway uses a self-created adaptive data compression and transmission algorithm to connect the data to the communication switch through Ethernet. The algorithm dynamically adjusts the compression ratio according to the network load and data importance. The formula is: Where C is the compression ratio, z i is the i-th important data item, t j is the jth non-important data item, k is the adjustment coefficient, N is the current network load, and T is the preset time period; the intelligent gateway and the process safety communication switch of the integrated cabinet of the sub-control center form a ring network structure to realize the transmission of the on-site collected data to the upper-level management unit; S3. Functional realization of interactive touch screen: The interactive touch screen is connected with the data acquisition unit and the storage and forwarding unit to obtain the system working status information in real time; The interactive touch screen provides a parameter setting interface for the data acquisition unit. The operator can enter the setting interface through specific touch gestures and password verification mechanisms to adjust the data acquisition frequency and warning threshold parameters. The parameter adjustment instructions are transmitted to the data acquisition unit through a custom secure communication protocol. The interactive touch screen has a data query function. It can retrieve and display the corresponding historical data from the data stored in the intelligent gateway according to the time range and data type query conditions entered by the operator. The display forms include charts and lists. The chart type is automatically adapted according to the characteristics of the data, including line charts to show the trend of water level changes over time and bar charts to show the statistics of valve opening and closing times; S4. Abnormal data detection and early warning: For abnormal data marked as to be detected, a detection method based on multivariate statistical analysis and machine learning models is used to select multiple relevant data variables, including valve action speed, water level change rate and equipment vibration frequency, to construct a data matrix. The principal component analysis (PCA) multivariate statistical analysis method is used to reduce the dimension of the data matrix and extract key features. The extracted features are input into the pre-trained support vector machine and neural network machine learning models for classification and judgment to determine whether the data is truly abnormal and the severity of the abnormality. If it is determined to be abnormal and the severity reaches the warning standard, a warning is issued in a variety of ways, including popping up a striking warning box on the interactive touch screen to display the abnormality details; At the same time, text messages and email notifications are sent to relevant managers. The warning information contains detailed analysis results of abnormal data, possible security risks and response suggestions. In addition, the collection parameters of the data collection unit are automatically adjusted according to the abnormal situation, including increasing the collection frequency and expanding the collection scope.

2. A method for process safety detection and early warning of a ship lock control system according to claim 1, characterized in that: In the CPU module of the data acquisition unit, the weight coefficient w in the safety assessment algorithm based on dynamic threshold and trend analysis is i and v j The method of determining is as follows: Regularly collect historical data of the ship lock under normal operation and simulated fault conditions, and pre-process these data, including removing outliers and normalizing the data; According to the valve opening data and the influence of different opening ranges on the safe operation of the ship lock, the weight coefficient w is determined by combining the analytic hierarchy process (AHP) and the fuzzy comprehensive evaluation method. i First, a judgment matrix is ​​constructed. The matrix elements are determined by expert experience and data analysis to reflect the relative importance of different opening data. Then, the judgment matrix is ​​fuzzified using the fuzzy comprehensive evaluation method to obtain a fuzzy relationship matrix. Finally, the weight vector w is obtained through fuzzy synthesis operation. i ; For the water level height data, according to the design water level standard of the ship lock and the actual operating water level fluctuation range, the weight coefficient is determined by combining the coefficient of variation method and the entropy weight method to calculate the coefficient of variation v of each water level data. j , reflecting the degree of data dispersion, and calculating the information entropy at the same time. The weight contribution of the data is determined according to the information entropy, and the weight coefficient v is obtained by combining the two. j .

3. A method for process safety detection and early warning of a ship lock control system according to claim 1, characterized in that: The digital input module in the data acquisition unit uses multi-sensor redundant acquisition technology when collecting data, specifically: Arrange multiple sensors of the same type at key locations, including multiple opening sensors at valves and multiple water level sensors at water level measurement points; Each sensor collects data independently and transmits the data to the data acquisition unit; The CPU module in the data acquisition unit uses a self-created algorithm based on data consistency check and weighted fusion to process the data collected by multiple sensors. The formula is: Where D is the fused data, d k is the data collected by the kth normal sensor, e l is the data collected by the lth suspected faulty sensor, a k and b l is the corresponding weighting coefficient, which is dynamically determined based on the historical reliability data of the sensor and the current data consistency. If the difference in data collected by multiple sensors exceeds the preset consistency threshold, it is judged that the sensor may be faulty, and its weighting coefficient is reduced and its data is directly excluded, and normal sensor data is used for fusion calculation.

4. A method for process safety detection and early warning of a ship lock control system according to claim 1, characterized in that: In the CAN bus communication protocol based on time division multiplexing and error correction coding adopted by the data communication module, the time slot allocation method of time division multiplexing is as follows: Dividing the communication time slots into different time slot groups according to the data types and transmission priorities collected by the data collection unit; For key data with high real-time requirements, including emergency valve closing instructions and fault alarm signals, more time slot resources are allocated for timely transmission; For common data, including equipment operation status monitoring data, relatively few time slot resources are allocated; The time slot allocation is dynamically adjusted according to the operating status of the lock. When the lock is in frequent operation or emergency, the time slot requirements of key data are prioritized. This dynamic time slot allocation method improves the communication efficiency of the CAN bus and the real-time performance of data transmission.

5. The method for process safety detection and early warning of a ship lock control system according to claim 1 is characterized in that: In the intelligent gateway of the storage and forwarding unit, the method for determining the adjustment coefficient in the adaptive data compression and transmission algorithm is as follows: Real-time monitoring of network performance indicators such as network bandwidth utilization, data transmission delay, and data loss rate; A comprehensive evaluation function is constructed based on these indicators. The function form is: E = α × U + β × D + γ × L, where E is the comprehensive evaluation value of network performance, U is the network bandwidth utilization, D is the data transmission delay, L is the data loss rate, and α, β, and γ are weight coefficients determined according to the network performance requirements of the ship lock control system; According to the range of the comprehensive evaluation value E, the value of the adjustment coefficient k is determined in the form of a piecewise function. When E is small, it indicates that the network performance is good. Increasing the k value appropriately improves the data compression ratio and reduces the data transmission volume. When E is large, it indicates that the network performance is poor. Reducing the k value reduces the compression ratio. In this way, the adjustment coefficient k is dynamically adjusted, so that the intelligent gateway can optimize the data compression and transmission strategy according to the actual network situation.

6. A method for process safety detection and early warning of a ship lock control system according to claim 1, characterized in that: The touch operation gesture recognition of the interactive touch screen adopts a gesture recognition model based on deep learning. The specific steps are as follows: Collect touch gesture samples of operators under different environmental conditions, including click, slide, and zoom gestures, and annotate the gesture samples, including gesture type, start position, end position, and duration information; Construct a convolutional neural network (CNN) model. The model structure includes convolutional layers, pooling layers, and fully connected layers. The convolutional layers are used to extract the features of gesture images, the pooling layers are used to reduce the data dimension, and the fully connected layers are used to classify gesture types. The CNN model is trained using the annotated gesture samples. During the training process, the stochastic gradient descent algorithm is used to optimize the model parameters. By continuously adjusting the parameters, the model can accurately recognize different touch operation gestures. The trained model is deployed on the interactive touch screen. When the operator performs a touch operation, the touch screen collects gesture images, which are input into the model for recognition after preprocessing. The corresponding operations are performed according to the recognition results, including opening the parameter setting interface and querying historical data.

7. A method for process safety detection and early warning of a ship lock control system according to claim 1, characterized in that: The secure communication protocol between the interactive touch screen and the data acquisition unit and the storage and forwarding unit adopts an encryption authentication mechanism based on elliptic curve encryption ECC and message authentication code MAC, specifically: Before data transmission, the sender uses the elliptic curve encryption algorithm to encrypt the transmitted data and generate ciphertext; At the same time, the sender uses the message authentication code algorithm to calculate the MAC value of the data and transmits the MAC value together with the ciphertext; After receiving the data, the receiver first uses the same elliptic curve encryption algorithm to decrypt the ciphertext to obtain the original data; Then, the receiver uses the same message authentication code algorithm to calculate the MAC value of the received data and compares it with the MAC value sent by the sender. If the two are consistent, it is considered that the data has not been tampered with and the source is reliable, and the data is accepted. Otherwise, the data is rejected and an alarm is issued.

8. The method for process safety detection and early warning of a ship lock control system according to claim 1 is characterized in that: When the digital output module in the data acquisition unit outputs the protection signal, it adopts a signal output strategy based on priority and delay time, specifically: Different priorities are set for different types of protection signals, including shutdown protection signal with the highest priority, emergency valve closing protection signal with the second highest priority, and other protection signals with lower priority; When multiple protection signals are generated simultaneously, they are output in order of priority; For each protection signal, a corresponding delay time is set according to the response characteristics and safety requirements of the ship lock equipment. If a higher priority protection signal is generated during the delay time, the output of the current protection signal is suspended and the higher priority protection signal is output first; The delay time is set based on the equipment dynamics model and safety time window analysis. By establishing a dynamic model of the lock equipment, analyzing the response time of the equipment under different working conditions, and combining the safety time window requirements to determine a reasonable delay time.

9. A method for process safety detection and early warning of a ship lock control system according to claim 1, characterized in that: In the intelligent gateway of the storage and forwarding unit, the hybrid storage structure based on data blocks and timestamps for data storage is specifically as follows: Divide the collected data into different data blocks according to data type and collection time, and each data block contains the same type of data within a certain time range; Add timestamp information to the header of each data block to record the start and end time of the data block; For frequently accessed data blocks, including recent key operating data, high-speed cache technology is used to store them to improve data query speed; For historical data blocks, hierarchical storage is performed based on the importance and access frequency of the data. Data blocks with high importance and high access frequency are stored in high-performance storage media, while other data blocks are stored in large-capacity, low-cost storage media. Create a data index table that records the storage location, data type, and time range information of the data block to facilitate data query and management.

Images