Ship lock intelligent full-automatic operation method and system based on Internet of Things

Abstract

The invention discloses a ship lock intelligent full-automatic operation method and system based on the Internet of Things, and relates to the technical field of water conservancy traffic, and the ship lock intelligent full-automatic operation method based on the Internet of Things comprises the steps of data collection, state analysis, dynamic decision making and remote optimization. By integrating the Internet of Things technology, the intelligent scheduling algorithm and the high-precision hydraulic synchronous calibration unit, full-automatic operation of the ship lock is achieved, and the operation efficiency and safety of a port are remarkably improved. The method ensures that ship berthing in the lock chamber meets the requirements of structural safety and navigation rules, effectively avoids the risk of intensive berthing, has important significance in the aspects of improving port operation efficiency, optimizing resource allocation, guaranteeing navigation safety and the like, and solves the problems of low efficiency and slow response caused by the fact that a traditional ship lock depends on manual operation. The method has the advantages of high reliability, low delay and expandability, and is suitable for inland river shipping and hydro-junction scenes.

Description

Technical Field

[0001] The present invention relates to the field of water conservancy and transportation technology, and in particular to an intelligent and fully automated operation method and system for a ship lock based on the Internet of Things. Background Art

[0002] Inland waterway transportation is a vital component of the nation's comprehensive transportation system and the integrated utilization of water resources. It is a key development direction for optimizing the transportation structure and building a green, efficient, and modern logistics system. Lock systems are key infrastructure in inland waterway transportation, and their operational efficiency and safety directly impact the smooth operation and profitability of inland waterway transportation. As inland waterway transportation enters a period of rapid development, the imbalance and inadequacy of the integrated utilization of inland water resources are becoming increasingly prominent.

[0003] Traditional ship lock operations rely on manual monitoring and control. Manual judgment of water level differences and ship positions is prone to errors, resulting in inappropriate timing for gate opening and closing. The ship locks respond slowly to emergencies such as sudden water level changes and equipment failures, posing safety hazards. The coordinated dispatch of ships is inefficient, impacting waterway capacity. Therefore, we propose an intelligent, fully automated operation method and system for ship locks based on the Internet of Things. Summary of the Invention

[0004] The purpose of the present invention is to solve the problems that traditional ship lock operations rely on manual monitoring and manual control, and manual judgment of water level differences and ship positions is prone to errors, resulting in inappropriate timing of gate opening and closing, slow response speed of the ship lock in emergencies such as sudden water level changes and equipment failures, safety hazards, low efficiency of ship coordinated scheduling, and impact on channel traffic capacity. The present invention provides an intelligent fully automated operation method and system for ship locks based on the Internet of Things.

[0005] In order to achieve the above-mentioned purpose, the present invention specifically adopts the following technical solutions:

[0006] An intelligent and fully automated operation method for a ship lock based on the Internet of Things, comprising:

[0007] Step 1) using a laser radar array, a strain sensor group, and a Beidou positioning terminal to collect ship three-dimensional profile data, gate stress distribution, and channel hydrological characteristic values in real time;

[0008] Step 2) A dynamic priority scheduling model is constructed. The ship load vector V = [W, D, C, R] is input, where W is the tonnage, D is the draft, C is the cargo type code, and R is the credit rating. The gear decision algorithm based on multi-objective optimization outputs the optimal gear plan. The priority scoring function is:

[0009] S=αW′+β(1-|D′-0.5|)+γ·C′ T u+δR′

[0010] Among them, α, β, γ, δ are adjustable weight coefficients, the default value is ∑=1, u=[u1,u2,…,u k ] is the urgency vector of cargo type, such as dangerous goods u i =1.5, ordinary goods u i =1.0,;

[0011] Objective function (minimize total waiting cost):

[0012]

[0013] Decision variable: x ijk ∈{0,1} indicates whether ship i is assigned to the kth berth of lock chamber j, is the expected waiting time of ship i, S i is the priority scoring function for ship i;

[0014] Constraints:

[0015] Unique location per vessel;

[0016] Lock chamber load limit;

[0017] Safety distance constraints;

[0018] Step 3) Solve the multi-objective optimization problem based on the improved NSGA-II algorithm, and use the adaptive crossover operator in the iterative process:

[0019]

[0020] Constraints: G current ∈[1,G max ];

[0021] Step 4) Send control instructions to the gate opening and closing machine through the industrial Internet of Things gateway to achieve millimeter-level synchronous control;

[0022] Step 5) The deformation trend of the door body is predicted based on the LSTM neural network. When the predicted displacement ΔX ≥ 3 mm, an early warning is triggered. The strain-displacement mapping model is:

[0023]

[0024] Where: ε i is the microstrain value of the i-th strain gauge, unit: με, k i is the position calibration coefficient, which is determined by calibration experiments and has an error of <0.1%. ambient is the ambient temperature, unit: °C;

[0025] Step 6) The sensor network collects water level, gate status, and ship position data in real time. The central control unit analyzes the data dynamically. When the water level change rate |ΔH / Δt| ≥ 0.5 m / h is detected, the emergency response mechanism is triggered. The actuator adjusts the gate status according to the control instructions and issues a navigation warning.

[0026] Furthermore, the calculation formula of the credit rating R in step 2) is:

[0027] R=0.6R history +0.3R emergency +0.1R penalty

[0028] in:

[0029] Furthermore, the step 3) adopts a space segmentation strategy to divide the lock chamber into n×m virtual grids and establish a two-dimensional discrete space coordinate system:

[0030]

[0031] Construct a binary state matrix P∈{0,1} n×m , whose elements satisfy:

[0032]

[0033] The maximum number of grid cells allowed to be occupied by the lock chamber is determined by structural safety and navigation regulations and must meet the following requirements:

[0034]

[0035] Among them, C max is the chamber capacity constant, usually satisfying C max <<n×m, to avoid the risk of dense parking.

[0036] Furthermore, the LSTM network in step 5) is:

[0037]

[0038] Input layer: 128-dimensional time series [ε_1,ε_2,...,ε_128], corresponding to a strain data sequence of 128 time steps;

[0039] Hidden layer: 256 GRU units;

[0040] Output layer: 3 warning levels [normal, caution, danger], using Softmax normalization.

[0041] Furthermore, the arrangement of the laser radar array in step 1) satisfies:

[0042]

[0043] Among them, D min is the minimum distance between adjacent radars, The maximum designed length of a navigable ship, in meters.

[0044] Furthermore, the water level sudden change response protocol in step 6) is: Level 1: tresponse≤30 seconds, Level 2: tresponse≤60 seconds, Level 3: tresponse≤90 seconds.

[0045] An intelligent and fully automated operation system for ship locks based on the Internet of Things, comprising:

[0046] An IoT data acquisition module, including a lidar array, a strain sensor group, and a BeiDou positioning terminal, is used to collect real-time lock operation data, including water level, gate status, vessel location, and environmental parameters;

[0047] The intelligent scheduling engine, connected to the IoT data acquisition module, is configured to analyze data and generate control instructions through a dynamic priority evaluation model and an improved NSGA-II algorithm;

[0048] Distributed control module, including PLC redundant controller and hydraulic synchronization calibration unit installed in electric valve, hydraulic drive device and alarm device, used to execute gate opening and closing, water level adjustment and abnormal state response according to control instructions;

[0049] The cloud management platform is connected to the central control unit through a wireless communication module to achieve remote monitoring and data storage.

[0050] Furthermore, the hydraulic synchronization calibration unit of the distributed control module includes a high-precision displacement sensor with a resolution of 0.01 mm and a proportional-integral controller. The transfer function of the proportional-integral controller is:

[0051]

[0052] Among them, T_f=0.1s is the filter time constant.

[0053] Furthermore, the data storage of the cloud management platform adopts a layered data structure, with the first layer storing device feature hash values, the second layer storing evidence scheduling decision logs, and the third layer recording emergency event handling processes.

[0054] Furthermore, the intelligent scheduling engine includes a ship behavior prediction subunit, which adopts Markov decision process modeling:

[0055] State S_t = [position, speed, cargo status]

[0056] Action A_t = {accelerate, hold, turn}

[0057] Reward R = on-time arrival reward - energy consumption penalty.

[0058] The beneficial effects of the present invention are as follows:

[0059] This invention integrates Internet of Things (IoT) technology, intelligent scheduling algorithms, and a high-precision hydraulic synchronization calibration unit to achieve fully automated operation of ship locks, significantly improving port operational efficiency and safety. The integrated application of a lidar array, strain sensor array, and Beidou positioning terminal enables real-time collection of ship 3D profile data, gate stress distribution, and channel hydrological characteristics, providing accurate data support for subsequent intelligent scheduling. A dynamic priority scheduling model comprehensively considers multiple factors, including ship tonnage, draft, cargo type, and credit rating, to calculate each ship's priority score in real time and output the optimal shifting plan to minimize total waiting costs. This model not only improves port throughput efficiency but also enables fairer and more efficient ship scheduling and optimizes port resource allocation. By solving multi-objective optimization problems using an improved NSGA-II algorithm and combining it with an adaptive crossover operator and a binary state matrix, this invention ensures that ships berthing within the lock chamber meet structural safety and navigation regulations while maintaining population diversity, effectively avoiding the risk of crowded berthing. This model is of great significance for improving port operational efficiency, optimizing resource allocation, and ensuring navigation safety. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below.

[0061] The present invention provides an intelligent and fully automated operation method for a ship lock based on the Internet of Things, comprising:

[0062] Step 1) using a laser radar array, a strain sensor group, and a Beidou positioning terminal to collect ship three-dimensional profile data, gate stress distribution, and channel hydrological characteristic values in real time;

[0063] The layout of the lidar array satisfies:

[0064]

[0065] Among them, D min is the minimum distance between adjacent radars, The maximum designed length of a navigable ship, in meters.

[0066] Step 2) A dynamic priority scheduling model is constructed. The ship load vector V = [W, D, C, R] is input, where W is the tonnage, D is the draft, C is the cargo type code, and R is the credit rating. The gear decision algorithm based on multi-objective optimization outputs the optimal gear plan. The priority scoring function is:

[0067] S=αW′+β(1-|D′-0.5|)+γ·C′ T u+δR′

[0068] Among them, α, β, γ, δ are adjustable weight coefficients, the default value is ∑=1, u=[u1,u2,…,u k ] is the urgency vector of cargo type, such as dangerous goods u i =1.5, ordinary goods u i =1.0,;

[0069] Objective function (minimize total waiting cost):

[0070]

[0071] Decision variable: x ijk ∈{0,1} indicates whether ship i is assigned to the kth berth of lock chamber j, is the expected waiting time of ship i, S i is the priority scoring function for ship i;

[0072] Constraints:

[0073] Unique location per vessel;

[0074] Lock chamber load limit;

[0075] Safety distance constraints;

[0076] Ship load vector V = [W, D, C, R]

[0077] W: Tonnage, which represents the ship's carrying capacity and is the main indicator of the amount of cargo transported by the ship.

[0078] D: Draft refers to the water depth of a ship when fully loaded. It is related to the ship's tonnage, ship type and cargo distribution, and affects whether the ship can safely pass through a specific channel or port.

[0079] C: Cargo type code, used to distinguish different types of cargo. Different types of cargo have different loading and unloading efficiency, storage requirements and transportation priorities.

[0080] R: Credit rating, which indicates the credit status of the ship or shipowner, which may affect the ship's scheduling priority, loading and unloading fees and port service priority.

[0081] Assume that the input vector V = [W, D, C, R] and normalize it according to the following rules:

[0082] Tonnage standardization

[0083] Draft safety factor

[0084] C′=OneHot(C), one-hot encoding of product type

[0085] Credit rating normalization

[0086] Among them D safe The maximum draft allowed for the lock chamber.

[0087]

[0088]

[0089] The dynamic priority scheduling model calculates each vessel's priority score in real time based on factors such as tonnage, draft, cargo type, and credit rating. Using a multi-objective optimization algorithm, the model outputs the optimal scheduling plan to minimize total waiting costs and improve port throughput efficiency. This model facilitates fairer and more efficient vessel scheduling, while also optimizing port resource allocation by factoring in safety and credit rating.

[0090] The calculation formula for credit rating R is:

[0091] R=0.6R history +0.3R emergency +0.1R penalty

[0092] in:

[0093] The calculation formula for credit rating R takes into account three factors: historical credit record, emergency dispatch performance and penalty record.

[0094] R: stands for credit rating, which is a comprehensive assessment result. The higher the value, the higher the credit rating.

[0095] R_history: indicates the score of the historical credit record. This score reflects the credit performance of the entity over the past period of time and is the basis for credit rating assessment.

[0096] R_emergency: This metric represents the score for emergency dispatch performance. The calculation formula is: number of successful dispatches / total number of dispatches × 100%. This metric measures the subject's ability to dispatch resources and resolve problems in emergency situations. A higher score indicates a better emergency dispatch performance.

[0097] R_penalty: This represents the penalty record score. This score reflects whether the subject has been penalized in the past and the severity of the penalty. The more and more severe the penalty record, the lower the score.

[0098] In terms of effectiveness, this formula comprehensively considers historical credit records, emergency dispatch performance, and penalty records, enabling a more comprehensive and objective assessment of an entity's credit rating. Furthermore, by adjusting the weighting of each component, it can emphasize or deemphasize certain aspects of the assessment as needed, making the credit rating assessment more tailored to actual needs.

[0099] Step 3) Solve the multi-objective optimization problem based on the improved NSGA-II algorithm, and use the adaptive crossover operator in the iterative process:

[0100]

[0101] Constraints: G current ∈[1,G max ];

[0102] P_crossover: stands for crossover probability, which is a parameter used in genetic algorithms. It determines the probability of using a crossover operation when selecting the parent generation for the crossover operation.

[0103] 0.9: is the initial value of the crossover probability, indicating that at the beginning of the genetic algorithm, the probability of the crossover operation is higher.

[0104] 0.5: is a coefficient used to adjust the rate at which the crossover probability changes over generations.

[0105] G_current: represents the current generation, indicating how many generations the genetic algorithm has evolved.

[0106] G_max: represents the maximum number of generations, which is a preset parameter of the genetic algorithm, indicating how many generations of evolution the genetic algorithm will undergo in total.

[0107] The constraint G_current∈[1,G_max] indicates that the value range of the current algebra G_current is from 1 to the maximum algebra G_max, which ensures the validity of the formula.

[0108] Based on the improved NSGA-II algorithm, a space partitioning strategy is adopted to solve the multi-objective optimization problem. The lock chamber is divided into n×m virtual grids and a two-dimensional discrete space coordinate system is established:

[0109]

[0110] Construct a binary state matrix P∈{0,1} n×m , whose elements satisfy:

[0111]

[0112] The maximum number of grid cells allowed to be occupied by the lock chamber is determined by structural safety and navigation regulations and must meet the following requirements:

[0113]

[0114] Among them, C max is the chamber capacity constant, usually satisfying C max <<n×m, to avoid the risk of dense parking.

[0115] A binary state matrix P∈{0,1}^n×m is constructed to describe the occupancy status of each grid cell within the lock chamber, where P_i,j=1 indicates that grid cell (i,j) is occupied by a vessel, and P_i,j=0 indicates that grid cell (i,j) is unoccupied. This step provides the foundation for subsequent ship scheduling and handling of lock chamber capacity constraints.

[0116] By calculating the total number of occupied grids in the lock chamber and comparing it with the maximum allowed total number of grids C_max in the lock chamber, it can be ensured that the berthing of ships in the lock chamber meets the requirements of structural safety and navigation regulations and avoids the risk of dense berthing.

[0117] Step 4) Send control instructions to the gate opening and closing machine through the Industrial Internet of Things gateway to achieve millimeter-level synchronous control; convert the optimization results into physical execution actions to ensure precise control of the gate.

[0118] Step 5) The deformation trend of the gate body is predicted based on the LSTM neural network. When the predicted displacement ΔX ≥ 3 mm, an early warning is triggered to predict the deformation trend of the gate structure and implement preventive maintenance. The strain-displacement mapping model is:

[0119]

[0120] Where: ε i is the microstrain value of the i-th strain gauge, unit: με, k i is the position calibration coefficient, which is determined by calibration experiments and has an error of <0.1%. ambient is the ambient temperature, unit: °C;

[0121] The strain-displacement mapping model converts the microstrain values of the strain gauge into corresponding displacement changes. The model considers the microstrain values of the strain gauge (εi, in με) and the corresponding position calibration coefficient (ki), which is determined through calibration experiments and has an error control of less than 0.1%. The model also accounts for the effect of ambient temperature (Tambient, in °C) by multiplying it by a fixed coefficient of 0.05.

[0122] This strain-displacement mapping model accurately calculates the displacement change (ΔX) by taking into account multiple factors, including strain value, position calibration coefficient, and ambient temperature. This mapping relationship is of great significance in fields such as structural health monitoring and material mechanical properties testing, enabling real-time monitoring and analysis of structural deformation and stress states.

[0123] The LSTM network is:

[0124]

[0125] ft = σ(Wf·ht-1,xt+bf): Forget Gate, where ft represents the output of the forget gate, σ represents the sigmoid function, Wf is the weight matrix of the forget gate, ht-1 represents the hidden state at the previous time step, xt represents the input at the current time step, and bf is the bias term of the forget gate. The forget gate determines which information in the cell state Ct-1 at the previous time step should be forgotten.

[0126] it = σ(Wi·ht-1,xt+bi): Input Gate, where it represents the output of the input gate, Wi is the weight matrix of the input gate, and bi is the bias term of the input gate. The input gate determines what new information in the input xt at the current time step needs to be added to the cell state Ct.

[0127] Ct = tanh(WC·ht-1,xt+bC): Candidate Cell State, where Ct represents the output of the candidate cell state, WC is the weight matrix of the candidate cell state, and bC is the bias term of the candidate cell state. The role of the candidate cell state is to generate a new candidate cell state value, which may be added to the current cell state.

[0128] Cell State Update, Represents element-by-element multiplication. This step applies the results of the forget gate and input gate to the cell state Ct-1 at the previous time step and the candidate cell state Ct at the current time step, thereby updating the cell state. The cell state is a key element in the LSTM network, used to preserve long-term dependency information.

[0129] ot = σ(Wo·ht-1,xt+bo): Output Gate, where ot represents the output of the output gate, Wo is the weight matrix of the output gate, and bo is the bias term of the output gate. The output gate determines which information in the cell state Ct is output to the hidden state ht at the current time step.

[0130] ht = ot°tanh(Ct): Hidden State Update. This step applies the output gate result to the current cell state Ct, which has been processed by the tanh activation function, to obtain the hidden state ht for the current time step. The hidden state is the output of the LSTM network, which is passed to the next time step or used for the final task (such as classification or regression).

[0131] Input layer: 128-dimensional time series [ε_1,ε_2,...,ε_128], corresponding to a strain data sequence of 128 time steps; the input layer is responsible for receiving time series data and passing it to the LSTM network or GRU network for processing

[0132] Hidden layer: 256 GRU units; The hidden layer consists of 256 GRU (Gated Recurrent Unit) units, which are used to process the time series data passed by the input layer and extract the feature information therein

[0133] Output layer: Three warning levels [Normal, Caution, Danger], using Softmax normalization. The output layer converts the hidden layer's output into the final warning level. Softmax is used for normalization, ensuring that the sum of the probabilities of the three output warning levels is 1. Ultimately, the network selects the warning level with the highest probability as the prediction based on the output probability distribution.

[0134] When ΔX ≥ 3mm, the three-level response is initiated:

[0135] ① Reduced load operation (opening and closing speed reduced to 50%);

[0136] ②Trigger structural health detection (ultrasonic flaw detection);

[0137] ③Generate a maintenance work order and push it to the management system.

[0138] Step 6) The sensor network collects water level, gate status, and ship position data in real time. The central control unit analyzes the data dynamically. When the water level change rate |ΔH / Δt| ≥ 0.5 m / h is detected, the emergency response mechanism is triggered. The actuator adjusts the gate status according to the control instructions and issues a navigation warning.

[0139] The water level sudden change response protocol is: Level 1: tresponse ≤ 30 seconds, Level 2: tresponse ≤ 60 seconds, Level 3: tresponse ≤ 90 seconds.

[0140] The working principle and usage process of the present invention: By integrating Internet of Things technology, intelligent scheduling algorithms and high-precision hydraulic synchronous calibration units, the fully automated operation of the ship lock is realized, which significantly improves the efficiency and safety of port operations. The integrated application of lidar arrays, strain sensor groups and Beidou positioning terminals makes it possible to collect the three-dimensional contour data of ships, gate stress distribution and channel hydrological characteristic values in real time, and provides accurate data support for subsequent intelligent scheduling. Through the dynamic priority scheduling model, multiple factors such as the tonnage, draft, cargo type and credit rating of the ship are comprehensively considered. The priority score of each ship can be calculated in real time, and the optimal gear scheme can be output to minimize the total waiting cost. This model not only improves the throughput efficiency of the port, but also realizes fairer and more efficient ship scheduling and optimizes the allocation of port resources. By solving the multi-objective optimization problem with an improved NSGA-II algorithm and combining the application of an adaptive crossover operator and a binary state matrix, the present invention ensures that the berthing of ships in the lock chamber meets the requirements of structural safety and navigation regulations while maintaining population diversity, effectively avoiding the risk of dense berthing. This is of great significance for improving port operation efficiency, optimizing resource allocation, and ensuring navigation safety.

[0141] An intelligent and fully automated operation system for ship locks based on the Internet of Things, comprising:

[0142] An IoT data acquisition module, including a lidar array, a strain sensor group, and a BeiDou positioning terminal, is used to collect real-time lock operation data, including water level, gate status, vessel location, and environmental parameters;

[0143] The intelligent scheduling engine, connected to the IoT data acquisition module, is configured to analyze data and generate control instructions through a dynamic priority evaluation model and an improved NSGA-II algorithm;

[0144] Distributed control module, including PLC redundant controller and hydraulic synchronization calibration unit installed in electric valve, hydraulic drive device and alarm device, used to execute gate opening and closing, water level adjustment and abnormal state response according to control instructions;

[0145] The cloud management platform is connected to the central control unit through a wireless communication module to achieve remote monitoring and data storage.

[0146] Furthermore, the hydraulic synchronization calibration unit of the distributed control module includes a high-precision displacement sensor with a resolution of 0.01 mm and a proportional-integral controller. The transfer function of the proportional-integral controller is:

[0147]

[0148] Among them, T_f=0.1s is the filter time constant. It is used to adjust the output according to the input signal in the hydraulic synchronous calibration unit to achieve precise position or speed control. G(s): The transfer function of the proportional-integral controller represents the conversion relationship from the input signal to the output signal. Kp: Proportional gain, which determines the response speed and amplitude of the controller to the input error. Increasing Kp can speed up the response, but may also cause system instability. Ki: Integral gain, used to eliminate static errors. The integral term accumulates past errors and adjusts the output accordingly to eliminate long-term errors. Kd represents the differential gain, which is used to predict future error changes and adjust the output in advance. s: Complex frequency variable, used to represent the conversion from the time domain to the frequency domain in the Laplace transform. Tf: Filter time constant, used to smooth the input signal or the output signal of the controller to reduce the impact of high-frequency noise on system performance. In the provided formula, Tf=0.1s, indicating that the time constant of the filter is 0.1 seconds.

[0149] The data storage of the cloud management platform adopts a hierarchical data structure. The first layer stores device feature hash values, the second layer stores scheduling decision logs, and the third layer records the emergency incident handling process. The design of this hierarchical data structure aims to improve the efficiency of data storage and retrieval while ensuring the integrity and traceability of the data. The device feature hash values stored in the first layer can quickly locate specific device data, facilitating subsequent analysis and processing. The scheduling decision logs stored in the second layer record the process and results of each scheduling decision in detail, providing strong support for subsequent emergency incident handling and data analysis. The emergency incident handling process recorded in the third layer can clearly show the entire process of emergency response, which helps to optimize the emergency response mechanism and improve the ability to respond to emergencies.

[0150] The intelligent scheduling engine includes a ship behavior prediction subunit, which adopts Markov decision process modeling:

[0151] State S_t = [position, speed, cargo status]. State S_t should contain the ship's current position, speed, and cargo status. Position can be latitude and longitude coordinates or relative to a point; speed is the ship's current speed; and cargo status can be full, partially loaded, or empty.

[0152] Action A_t = {accelerate, hold, turn}. Action A_t should include acceleration, holding the current speed, and turning. Acceleration can mean increasing the speed by a certain amount; holding means maintaining the current speed; and turning means adjusting the course left or right.

[0153] Reward R = On-time arrival reward - Energy consumption penalty. The reward function R should include both the on-time arrival reward and the energy consumption penalty. If a ship reaches its destination within the scheduled time, a certain reward is given; at the same time, a penalty is applied based on the ship's energy consumption, with higher energy consumption resulting in greater penalties.

[0154] Using the state space, action space, and reward function defined above, a Markov decision process model is constructed. The model learns which actions to take in different states to maximize long-term rewards. By predicting ship behavior, the intelligent scheduling engine can more effectively plan ship routes, reducing waiting time and navigation conflicts, thereby improving overall scheduling efficiency. By considering energy consumption penalties, the intelligent scheduling engine can guide ships to adopt more energy-efficient navigation strategies, thereby reducing operating costs and benefiting environmental protection. The intelligent scheduling engine can monitor the status and behavior of ships in real time, promptly identifying and warning of potential safety hazards, and improving navigation safety and reliability. By optimizing scheduling strategies, the intelligent scheduling engine can ensure that ships arrive at their destinations on time, improving user experience and satisfaction.

[0155] The working principle and usage process of the present invention are as follows: When the system is in use, the IoT data acquisition module continuously collects lock operation data, including information such as water level changes, gate opening and closing status, real-time location of ships, and environmental parameters. This data is transmitted to the intelligent scheduling engine in real time via a high-speed communication link. After receiving the data, the intelligent scheduling engine immediately activates a dynamic priority assessment model to prioritize the passage requirements of each ship. At the same time, the improved NSGA-II algorithm generates a series of feasible scheduling solutions based on the current operating status of the lock, the priority scores of the ships, and possible conflicts. The algorithm continuously iterates and optimizes until the optimal solution, i.e., the optimal scheduling solution, is found. Once the optimal scheduling solution is determined, the intelligent scheduling engine sends control instructions to actuators such as electric valves and hydraulic drive devices through the distributed control module. The PLC redundant controller in the distributed control module ensures the accurate execution of the instructions. At the same time, the hydraulic synchronization calibration unit uses feedback from the high-precision displacement sensor and proportional-integral controller to achieve precise control of actions such as gate opening and closing and water level adjustment.

[0156] Throughout operation, the cloud-based management platform receives and stores real-time data from the IoT data collection module and the intelligent scheduling engine, including information such as equipment status, scheduling decision logs, and emergency response procedures. This data is stored in a hierarchical data structure, facilitating subsequent data analysis and traceability. The ship behavior prediction subunit within the intelligent scheduling engine utilizes a Markov decision process model to predict future ship behavior. By predicting information such as a ship's position, speed, and cargo status, the intelligent scheduling engine can more effectively plan ship routes, reducing waiting times and navigation conflicts, thereby improving overall scheduling efficiency. Through real-time data collection, intelligent scheduling, precise control, and remote monitoring, the intelligent and fully automated operation of the locks is achieved. This not only improves the efficiency and safety of the locks but also reduces operating costs, providing strong support for the intelligent development of waterway transportation.

[0157] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for intelligent and fully automated operation of ship locks based on the Internet of Things, characterized in that: include: Step 1) using a laser radar array, a strain sensor group, and a Beidou positioning terminal to collect ship three-dimensional profile data, gate stress distribution, and channel hydrological characteristic values in real time; Step 2) A dynamic priority scheduling model is constructed. The ship load vector V = [W, D, C, R] is input, where W is the tonnage, D is the draft, C is the cargo type code, and R is the credit rating. The gear decision algorithm based on multi-objective optimization outputs the optimal gear plan. The priority scoring function is: S=αW′+β(1-|D′-0.5|)+γ·C′ T u+δR′ Among them, α, β, γ, δ are adjustable weight coefficients, the default value is ∑=1, u=[u1,u2,…,u k ] is the urgency vector of cargo type, such as dangerous goods u i =1.5, ordinary goods u i =1.0,; Objective function (minimize total waiting cost): Decision variable: x ijk ∈{0,1} indicates whether ship i is assigned to the kth berth of lock chamber j, is the expected waiting time of ship i, S i is the priority scoring function for ship i; Constraints: Unique location per vessel; Lock chamber load limit; Safety distance constraints; Step 3) Solve the multi-objective optimization problem based on the improved NSGA-II algorithm, and use the adaptive crossover operator in the iterative process: Constraints: G current ∈[1,G max ]; Step 4) Send control instructions to the gate opening and closing machine through the industrial Internet of Things gateway to achieve millimeter-level synchronous control; Step 5) The deformation trend of the door body is predicted based on the LSTM neural network. When the predicted displacement ΔX ≥ 3 mm, an early warning is triggered. The strain-displacement mapping model is: Where: ε i is the microstrain value of the i-th strain gauge, unit: με, k i is the position calibration coefficient, which is determined by calibration experiments and has an error of <0.1%. ambient is the ambient temperature, unit: °C; Step 6) The sensor network collects water level, gate status, and ship position data in real time. The central control unit analyzes the data dynamically. When the water level change rate |ΔH / Δt| ≥ 0.5 m / h is detected, the emergency response mechanism is triggered. The actuator adjusts the gate status according to the control instructions and issues a navigation warning.

2. The method for intelligent and fully automated operation of a ship lock based on the Internet of Things according to claim 1, characterized in that: The calculation formula of the credit rating R in step 2) is: <h2 style=";text-align:left;direction:ltr">R=0.6R<h2 style=";text-align:left;direction:ltr"> history <h2 style=";text-align:left;direction:ltr"> +0.3R<h2 style=";text-align:left;direction:ltr"> emergency <h2 style=";text-align:left;direction:ltr"> +0.1R<h2 style=";text-align:left;direction:ltr"> penalty in:

3. The method for intelligent and fully automated operation of a ship lock based on the Internet of Things according to claim 1, characterized in that: The step 3) adopts a space segmentation strategy to divide the lock chamber into n×m virtual grids and establish a two-dimensional discrete space coordinate system: Construct a binary state matrix P∈{0,1} n×m , whose elements satisfy: The maximum number of grid cells allowed to be occupied by the lock chamber is determined by structural safety and navigation regulations and must meet the following requirements: Among them, C max is the chamber capacity constant, usually satisfying C max <<n×m, to avoid the risk of dense parking.

4. The method for intelligent and fully automated operation of a ship lock based on the Internet of Things according to claim 1, characterized in that: The LSTM network in step 5) is: Input layer: 128-dimensional time series [ε_1,ε_2,...,ε_128], corresponding to a strain data sequence of 128 time steps; Hidden layer: 256 GRU units; Output layer: 3 warning levels [normal, caution, danger], using Softmax normalization.

5. The method for intelligent and fully automated operation of a ship lock based on the Internet of Things according to claim 1, characterized in that: The arrangement of the laser radar array in step 1) satisfies: Among them, D min is the minimum distance between adjacent radars, The maximum designed length of a navigable ship, in meters.

6. The method for intelligent and fully automated operation of a ship lock based on the Internet of Things according to claim 1, characterized in that: The water level sudden change response protocol in step 6) is: Level 1: tresponse≤30, Level 2: tresponse ≤ 60 seconds, Level 3: tresponse ≤ 90 seconds.

7. An intelligent fully automated operation system for ship locks based on the Internet of Things, characterized by: include: An IoT data acquisition module, including a lidar array, a strain sensor group, and a BeiDou positioning terminal, is used to collect real-time lock operation data, including water level, gate status, vessel location, and environmental parameters; The intelligent scheduling engine, connected to the IoT data acquisition module, is configured to analyze data and generate control instructions through a dynamic priority evaluation model and an improved NSGA-II algorithm; Distributed control module, including PLC redundant controller and hydraulic synchronization calibration unit installed in electric valve, hydraulic drive device and alarm device, used to execute gate opening and closing, water level adjustment and abnormal state response according to control instructions; The cloud management platform is connected to the central control unit through a wireless communication module to achieve remote monitoring and data storage.

8. The intelligent fully automated operation system for ship locks based on the Internet of Things according to claim 7 is characterized by: The hydraulic synchronization calibration unit of the distributed control module includes a high-precision displacement sensor with a resolution of 0.01 mm and a proportional-integral controller. The transfer function of the proportional-integral controller is: Among them, T_f=0.1s is the filter time constant.

9. The intelligent fully automated operation system for ship locks based on the Internet of Things according to claim 7, characterized in that: The data storage of the cloud management platform adopts a layered data structure, the first layer stores device feature hash values, the second layer stores scheduling decision logs, and the third layer records emergency event handling processes.

10. The intelligent fully automated operation system for ship locks based on the Internet of Things according to claim 7, characterized in that: The intelligent scheduling engine includes a ship behavior prediction subunit, which adopts Markov decision process modeling: State S_t = [position, speed, cargo status] Action A_t = {accelerate, hold, turn} Reward R = on-time arrival reward - energy consumption penalty.

Images