Ship engine control system

Abstract

The invention discloses a ship engine control system, and relates to the technical field of engine control. A ship engine control system comprises a data processing module, a data prediction module, a decision optimization module, a parameter generation module and a control execution module. According to the invention, a plurality of disturbance prediction parameters such as surge arrival time, environment temperature change, port congestion, pitching and rolling change trend, wave period and countercurrent intensity change trend are fused in the data prediction module, and a disturbance arrival time window and an intensity level are calculated, so that potential navigation disturbance is identified in advance; and a basis is provided for feed-forward adjustment of power distribution and combustion control, so that active intervention is realized before disturbance occurs.

Description

Technical Field

[0001] The present invention relates to the technical field of engine control, and in particular to a ship engine control system. Background Art

[0002] During navigation, the operating status and energy management of a ship's propulsion system directly impact its safety, economy, and environmental adaptability. With the increasing complexity of ship propulsion systems, especially the increasing use of hybrid systems combining traditional fuel and electric propulsion, achieving efficient and stable power control under variable sea conditions, complex navigation missions, and multiple energy supply conditions has become a key research direction in ship power control technology.

[0003] Existing ship engine control systems usually perform engine load regulation and combustion control based on a single or small amount of sensor data. They lack real-time assessment and credibility verification of the health status of sensor data, and are prone to a decrease in control accuracy due to sensor drift, failure, or sudden changes in sea conditions. In addition, existing systems often use fixed strategies for combustion and propulsion power distribution, which makes it difficult to respond to disturbances such as surges, backflows, pitch and roll in a timely manner. This results in transient mechanical shocks in the propulsion system and engine under severe sea conditions, increasing component wear and energy consumption. When sailing at sea, ships face complex working conditions such as sudden changes in ambient temperature, changes in wave cycle, and port congestion. These factors will directly affect fuel atomization, combustion efficiency, propeller load changes, and power system balance. Existing technologies lack the ability to fuse and predict multi-source data for these working conditions, and are unable to perform power curve smoothing and control parameter feedforward correction in advance, resulting in insufficient transition control before the arrival of disturbances.

[0004] Furthermore, existing multi-objective optimization methods for ship propulsion control remain inadequate. For one thing, optimization often focuses on a single performance metric, failing to comprehensively consider multiple objectives, such as speed, navigation stability, range mission completion, and energy consumption. Furthermore, optimization calculations often fail to consider the dynamic response characteristics of engines and propulsion actuators, power rate limits, and safety margins in harsh environments. This makes it difficult to balance performance and stability in actual operation. Summary of the Invention

[0005] The present invention proposes a ship engine control system that realizes multi-source data health monitoring and trustworthy fusion, has disturbance prediction and feedforward shaping capabilities, and dynamically generates control information under multi-objective optimization constraints, so as to improve the energy efficiency, stability and reliability of ships under complex and changeable sea conditions and mission conditions.

[0006] A ship engine control system, comprising: Data processing module: acquires multi-source data of the ship during navigation, performs health monitoring, trust analysis and weighted fusion processing on it, and obtains fused navigation status information; Data prediction module: obtains expected navigation status information and expected demand information in sequence; Decision Optimization Module: Based on the integration of navigation status information, expected navigation status information and expected demand information, a multi-objective optimization model including speed, navigation stability, range mission and energy consumption is established. The power change smoothing constraint and impact cost weight are introduced within the disturbance arrival time window contained in the expected navigation status information to determine the ship's expected decision. Parameter Generation Module: Generates control information, including combustion control parameters, propulsion control parameters, and electrical management parameters, based on expected navigation state information and expected decisions. Within the disturbance arrival time window contained in the expected navigation state information, the control information is feedforward shaped and slope-limited according to the intensity level. When a severe environmental mode is detected, a weight matrix corresponding to the type and level is obtained to perform a weighted update on the control information. Control execution module: After receiving the control information output by the parameter generation module, it combines the disturbance arrival time window and intensity level contained in the expected navigation status information, as well as the dynamic response characteristics of the propulsion and combustion actuators, to make real-time fine-tuning of the combustion control parameters and propulsion power distribution, and updates the engine control execution instructions based on the fine-tuned control information.

[0007] As a preferred technical solution of the present invention, obtaining the fused navigation status information includes: Perform health monitoring on each source device of data acquired through sensors in multi-source data to generate a corresponding health status; when the health status of a source device is abnormal, trigger virtual compensation to generate alternative data; the health monitoring includes detecting data drift, signal fluctuation amplitude, abnormal delay and data loss; Conduct trustworthy analysis on multi-source data, including calculating the health metrics between similar data based on the health status of different source devices of the same data; verify the association of different types of data through the data association relationship table, calculate the association metrics between different types of data, and calculate the credibility factor of each type of data based on the weighted health metrics and association metrics; obtain the corresponding credibility weight based on the credibility factor, and perform weighted fusion processing on the multi-source data to obtain fused navigation status information.

[0008] As a preferred technical solution of the present invention, the virtual compensation includes: Obtain the health status of the remaining source devices of the same type as the target source device. When the proportion of devices with a healthy health status is higher than a first threshold and the proportion of devices with an abnormal health status is lower than a second threshold, perform fitting estimation based on the data corresponding to the healthy devices to generate alternative data corresponding to the current source device; otherwise, call the estimation model to perform inference based on historical data and relevant data in the data association relationship table. The estimation model is a time series prediction model based on machine learning.

[0009] As a preferred technical solution of the present invention, the sequential acquisition of expected navigation status information and expected demand information includes: performing navigation prediction based on the fused navigation status information, including surge arrival time prediction, ambient temperature change prediction, port congestion prediction, pitch angle and roll angle change trend prediction, wave period prediction and backflow intensity change trend prediction, and calculating the disturbance arrival time window and intensity level accordingly, generating expected navigation status information including the above-mentioned prediction parameters; based on the fused navigation status information and expected navigation status information, combined with historical navigation path data and mission requirement path data, performing speed requirement and stability requirement prediction to generate expected demand information.

[0010] As a preferred technical solution of the present invention, the multi-objective optimization model parameterizes the control variables during the operation of the ship into expected decisions including target track, target output power, idle condition maintenance time, combustion mode switching strategy, propulsion power allocation strategy, propeller pitch change strategy and electric propulsion power allocation strategy, and constructs a cost function in the multi-objective optimization model with speed, navigation stability, range mission and energy consumption as comprehensive optimization objectives; the cost function is dynamically calculated based on the real-time navigation parameters in the fused navigation status information, the disturbance arrival time window and intensity level parameters contained in the expected navigation status information, and the expected demand information.

[0011] As a preferred technical solution of the present invention, the dynamic calculation includes: introducing a dynamic response model of the propulsion and combustion actuators, power output upper limit and change rate limit, fuel and power supply capacity boundary, structural load and vibration threshold and emission regulation limit; and applying power change smoothing constraint within the disturbance arrival time window to limit the change rate of the target power curve, and at the same time introducing impact cost weight associated with the disturbance intensity level to reduce the transient mechanical shock of the propulsion system and the engine, and obtaining the optimization solution that minimizes the comprehensive cost function under the above constraints through a multi-objective optimization solution method, and outputting the optimization solution to the parameter generation module.

[0012] As a preferred technical solution of the present invention, the generation of control information includes: in the process of generating combustion control parameters, using the ambient temperature change prediction in the expected navigation status information to correct the combustion mode switching strategy in the expected decision; based on the disturbance prediction in the expected navigation status information and its arrival time window and intensity level, the target output power curve in the expected decision is feedforward adjusted; in the process of generating propulsion control parameters, using the wave period prediction and the backflow intensity change trend prediction to optimize the propulsion power distribution and propeller pitch change strategy in the expected decision; in the process of generating electrical management parameters, using the port congestion prediction to adjust the electric propulsion power distribution strategy in the expected decision.

[0013] As a preferred technical solution of the present invention, the feedforward shaping and slope limiting processing includes: within a preset advance amount of the disturbance arrival time window, using the disturbance intensity level to determine the allowable variation range and change rate of the injection quantity, rail pressure and air-fuel ratio in the combustion control parameters, applying slope limitations to the torque command and propeller pitch command in the propulsion control parameters, and adjusting the control parameter curve through a smooth transition function.

[0014] As a preferred technical solution of the present invention, the weighted update includes: monitoring operating parameters including ambient temperature, sea condition level, port congestion and mission type by fusing navigation status information with expected navigation status information; when a single operating condition parameter or a combination of parameters exceeds the corresponding preset threshold, it is determined to enter a harsh environment mode, and the type and level of the harsh environment mode are determined according to the type and amplitude of the exceeded parameter; obtaining a weight matrix corresponding to the type and level, mapping the matrix elements to each control variable of the combustion control parameter, propulsion control parameter and electrical management parameter, applying a proportional correction coefficient to the combustion control parameter to improve combustion stability or fuel atomization effect, applying a proportional correction coefficient to the propulsion control parameter to increase the power stability margin or limit the propeller pitch change rate, and applying a proportional correction coefficient to the electrical management parameter to optimize the electric propulsion power distribution or adjust the load switching sequence.

[0015] As a preferred technical solution of the present invention, the real-time fine-tuning includes: Based on the disturbance arrival time window and intensity level, combined with the dynamic response characteristics of the propulsion and combustion actuators, the preset advance amount before the disturbance arrives and the corresponding adjustment amplitude are calculated; in the real-time fine-tuning of the combustion control parameters, the instantaneous change amplitude of the injection quantity, rail pressure and air-fuel ratio are adjusted according to the disturbance intensity level, and a gradual correction is implemented before the disturbance arrives; in the real-time fine-tuning of the propulsion power distribution, the change rate of the torque command and the propeller pitch command is limited according to the dynamic response characteristics of the actuator, and a smooth output is maintained during the disturbance effect. The execution instructions of the engine control are synchronously updated according to the control information after fine-tuning, so as to maintain the optimized state of navigation stability and energy efficiency before, during and after the disturbance arrives.

[0016] The present invention has the following advantages: The present invention introduces multi-source data health monitoring, trustworthy analysis and weighted fusion processing into the data processing module. In the event of sensor drift, delay, fluctuation or failure, alternative data is generated through virtual compensation to ensure the continuity and reliability of input data, providing highly reliable fused navigation status information for subsequent predictions and decision-making.

[0017] The present invention integrates multiple disturbance prediction parameters such as surge arrival time, ambient temperature changes, port congestion, pitch and roll change trends, wave period and backflow intensity change trends into the data prediction module, and calculates the disturbance arrival time window and intensity level. It can identify potential navigation disturbances in advance, provide a basis for feedforward adjustment of power allocation and combustion control, and thus achieve active intervention before the disturbance occurs.

[0018] The present invention constructs a multi-objective optimization model with speed, navigation stability, range mission and energy consumption as comprehensive optimization objectives in the decision optimization module, and introduces the dynamic response model of propulsion and combustion actuators, power change smoothing constraints and impact cost weights to generate expected decisions that take into account performance, stability and equipment life under complex sea conditions and mission conditions.

[0019] The present invention utilizes disturbance prediction parameters in a parameter generation module to perform feedforward shaping and slope limiting on the combustion mode switching strategy, propulsion power distribution strategy, propeller pitch change strategy, and electric propulsion power distribution strategy, and updates the control parameters based on the type and level weight matrix in harsh environment modes, thereby effectively reducing mechanical shock, improving combustion efficiency, and enhancing the environmental adaptability of the system.

[0020] The present invention combines the disturbance arrival time window and intensity level, as well as the dynamic response characteristics of the actuator in the control execution module, to perform real-time fine-tuning of combustion control parameters and propulsion power distribution, and dynamically optimizes engine control instructions before, during, and after the disturbance, significantly improving the ship's navigation stability, propulsion system life, and overall energy efficiency in complex sea conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order 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 schematic diagrams of the present invention. Those skilled in the art can also derive other drawings based on the provided drawings without inventive effort. Figure 1 This is a schematic structural diagram of a ship engine control system used in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] To make the objectives, technical solutions, and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings. 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.

[0023] Embodiment, a ship engine control system, see Figure 1 As shown, it includes the following modules: Data processing module: acquires multi-source data of the ship during navigation, performs health monitoring, trust analysis and weighted fusion processing on it, and obtains fused navigation status information; Obtaining the fused navigation status information includes: Perform health monitoring on each source device of data acquired through sensors in multi-source data to generate a corresponding health status; when the health status of a source device is abnormal, trigger virtual compensation to generate alternative data; the health monitoring includes detecting data drift, signal fluctuation amplitude, abnormal delay and data loss; Conduct trustworthy analysis on multi-source data, including calculating the health metrics between similar data based on the health status of different source devices of the same data; verify the association of different types of data through the data association relationship table, calculate the association metrics between different types of data, and calculate the credibility factor of each type of data based on the weighted health metrics and association metrics; obtain the corresponding credibility weight based on the credibility factor, and perform weighted fusion processing on the multi-source data to obtain fused navigation status information.

[0024] The virtual compensation includes: Obtain the health status of the remaining source devices of the same type as the target source device. When the proportion of devices with a healthy health status is higher than a first threshold and the proportion of devices with an abnormal health status is lower than a second threshold, perform fitting estimation based on the data corresponding to the healthy devices to generate alternative data corresponding to the current source device; otherwise, call the estimation model to perform inference based on historical data and relevant data in the data association relationship table. The estimation model is a time series prediction model based on machine learning.

[0025] In one embodiment, the multi-source data includes but is not limited to the following categories: Hull attitude and motion data: acquired through onboard sensors such as the inertial measurement unit (IMU), gyroscope, accelerometer, and ship speed logger, including pitch angle, roll angle, bow angle, heave rate, longitudinal acceleration, etc., with an update frequency of 10-100 Hz to ensure dynamic response accuracy.

[0026] Propulsion and combustion system operating parameters: from the main engine control unit (ECU) and the propulsion system monitoring device, including main engine speed, torque, fuel injection amount, combustion chamber pressure, rail pressure, air-fuel ratio, propeller pitch angle and pitch rate, etc. The data collection period is set to 1 to 5 seconds.

[0027] External environment and sea condition information: collected through meteorological sensors (wind speed and direction meter, temperature and humidity meter), wave radar, sonar depth sounder, etc., including wind speed and direction, ambient temperature, sea water temperature, wave height, wave period, countercurrent intensity, etc.; some data are obtained from the shore-based meteorological service platform through the ship-shore satellite communication interface.

[0028] Navigation and positioning information: provided by the ship's GNSS system (GPS / Beidou), heading indicator, speedometer, etc., including the ship's real-time position, heading, speed, track deviation, etc.

[0029] Port and waterway traffic information: received through AIS (Automatic Identification System), VTS (Vessel Traffic Service System), etc., including dynamic information of nearby ships, port berth conditions, waterway congestion index, etc.

[0030] Obtaining the fused navigation status information includes: Health monitoring: Real-time health status detection is performed on the output signal of each source device in the multi-source data. The health status can be divided into "healthy", "slight deviation" and "abnormal". Abnormal status includes: data drift, monitoring the trend of output values ​​deviating from the historical statistical mean for a long time. Abnormal signal fluctuation amplitude, monitoring the situation where the signal change in a short period of time exceeds the physically possible range. Abnormal delay, the signal transmission delay exceeds the upper limit set by the system, such as 200ms. Data missing, no new data packet is received during the sampling period. Strong swing disturbance, judged when the pitch or roll angle change rate exceeds the preset safety threshold. Equipment failure, triggered when no signal is received for multiple consecutive periods or the data remains unchanged. When the health status is judged to be abnormal, virtual compensation is triggered to generate alternative data to ensure data continuity and reliability.

[0031] The first preferred compensation method is to obtain signals from other healthy devices of the same type as the target source device and generate compensation values ​​through weighted fitting or interpolation. For example, if the fuel flow meter on host 1 is abnormal, the compensation value can be calculated based on the fuel flow meter signal on host 2 and the load distribution ratio.

[0032] The second preferred compensation method: If the proportion of healthy devices is lower than the set threshold, the time series prediction model based on machine learning is called to combine historical data and data association table to infer the current data value.

[0033] Credibility Analysis: Cross-validation and weighting of multi-source data. Based on health monitoring results, a health score is calculated for each type of data (e.g., multiple speed sensor outputs). Consistency scores for different data types are calculated based on a data association table (pre-established physical coupling models, such as the relationship between propulsion power and ship speed). A weighted combination of the health score and consistency score is used to determine a credibility factor (ranging from 0 to 1) for each data type.

[0034] Dynamic weights are assigned to each data source according to the credibility factor. Multi-source heterogeneous data are fused into unified format fused navigation status information through the Bayesian fusion method, and used as the input of the subsequent data prediction module.

[0035] Data prediction module: obtains expected navigation status information and expected demand information in sequence; The sequential acquisition of expected navigation status information and expected demand information includes: performing navigation prediction based on the fused navigation status information, including surge arrival time prediction, ambient temperature change prediction, port congestion prediction, pitch angle and roll angle change trend prediction, wave period prediction and countercurrent intensity change trend prediction, and calculating the disturbance arrival time window and intensity level accordingly, and generating expected navigation status information including the above-mentioned prediction parameters; based on the fused navigation status information and expected navigation status information, combined with historical navigation path data and mission requirement path data, performing speed requirement and stability requirement prediction to generate expected demand information.

[0036] In one embodiment, the data prediction module completes the forward-looking prediction of navigation status and mission requirements based on the fusion of navigation status information, combined with external real-time data and historical data models, and provides input basis for the decision optimization module.

[0037] Power variation smoothing constraint: An optimization constraint that limits the power variation rate to prevent mechanical stress concentration caused by large power adjustments in a short period of time.

[0038] Impact cost weight: A penalty coefficient set based on the disturbance intensity level, used to additionally suppress the transient impact effect caused by the disturbance during the optimization process.

[0039] Obtaining the expected navigation status information and expected demand information in sequence includes: Sailing prediction based on integrated navigation status information includes, but is not limited to, the following sub-tasks: Surge arrival time prediction: This uses ship attitude changes, wave radar measurement data, and meteorological station wind and wave forecasts, combined with a wave propagation velocity model, to calculate the timestamp of the surge's arrival at the ship's location within a certain timeframe. Ambient temperature change prediction: This combines current ambient temperature, seawater temperature, weather trend forecasts, and the ship's speed and course to estimate the temperature curve for the future voyage. This prediction directly impacts the air-fuel ratio setting and fuel atomization performance in the combustion control strategy. Port congestion prediction: This uses AIS / VTS data, combined with port berth scheduling information and historical traffic flow curves, to calculate berth occupancy and queue waiting time upon arrival at the port. Pitch and roll angle change trend prediction: This uses historical ship attitude data, wave period, and wave direction information to predict future roll angle change trends and amplitudes through frequency domain analysis (FFT) and state-space models. Wave period prediction: This uses wave radar and historical sea condition data to obtain wave period trends through spectral analysis. This prediction is used to optimize the power regulation rhythm. The trend of changes in the intensity of the countercurrent is predicted by combining the ship speed log, GNSS ground speed difference and ocean current monitoring buoy data to establish a countercurrent change trend model and predict the changes in flow speed in future sections.

[0040] After completing the above predictions, the system calculates the disturbance arrival time window (i.e., the time range within which external disturbances affect the ship) and intensity level (disturbance impact level classification, such as 1 to 5) based on the prediction results, and integrates all prediction parameters to generate expected navigation status information.

[0041] Based on the fused navigation state information and expected navigation state information, combined with historical navigation path data (such as navigation parameters for the same route over the past five years) and mission requirement path data (such as planned arrival time, cargo loading and unloading time windows, and energy consumption control targets), the following predictions are performed: Speed ​​requirement prediction: Based on the current speed, remaining range, and tide and current forecasts, the optimal speed range that meets the mission time constraints is calculated. Stability requirement prediction: Combining data such as the expected pitch and roll predictions and wave period predictions from the expected navigation state, the attitude control target value that meets navigation comfort and safety requirements is calculated. Energy consumption and emission constraint prediction: Based on emission regulations, fuel inventory, power reserves, and refueling plans, the fuel and power consumption allocation strategy for future legs is calculated. Ultimately, expected demand information is generated and used as one of the inputs to the decision-making optimization module. Together with the fused navigation state information and the expected navigation state information, it drives the multi-objective optimization process.

[0042] The disturbance arrival time window refers to the time range within which external disturbances (such as large waves, strong winds, and headwinds) have a significant impact on ships, and is calculated comprehensively by multiple prediction models.

[0043] Intensity level, a quantitative level of the impact of external disturbances on the ship's power system, attitude stability and fuel economy, which can be calculated based on parameters such as wave height, wind speed, and current velocity.

[0044] Decision Optimization Module: Based on the integration of navigation status information, expected navigation status information and expected demand information, a multi-objective optimization model including speed, navigation stability, range mission and energy consumption is established. The power change smoothing constraint and impact cost weight are introduced within the disturbance arrival time window contained in the expected navigation status information to determine the ship's expected decision. The multi-objective optimization model parameterizes the control variables during the ship operation process into expected decisions including target track, target output power, idle condition maintenance time, combustion mode switching strategy, propulsion power allocation strategy, propeller pitch change strategy and electric propulsion power allocation strategy, and constructs a cost function in the multi-objective optimization model with speed, navigation stability, range mission and energy consumption as comprehensive optimization objectives; the cost function is dynamically calculated based on the real-time navigation parameters in the fused navigation status information, the disturbance arrival time window and intensity level parameters contained in the expected navigation status information, and the expected demand information.

[0045] The dynamic calculation includes: introducing dynamic response models of propulsion and combustion actuators, power output upper limits and change rate limits, fuel and power supply capacity boundaries, structural load and vibration thresholds, and emission regulations limits; and applying power change smoothing constraints within the disturbance arrival time window to limit the change rate of the target power curve. At the same time, an impact cost weight associated with the disturbance intensity level is introduced to reduce the transient mechanical impact of the propulsion system and engine. An optimization solution that minimizes the comprehensive cost function under the above constraints is obtained through a multi-objective optimization solution method, and the optimization solution is output to a parameter generation module.

[0046] In one embodiment, the multi-objective optimization model parameterizes the control variables during the operation of the ship into the following: target track, which is the optimal route calculated by the path planning algorithm under the conditions of obstacle avoidance, safety distance and navigation regulations. Target output power, which corresponds to the power setting value of the engine and propulsion system in different sections, is constrained by energy consumption optimization and propulsion stability requirements. Idle condition holding time, which is the optimized value of the engine idling duration in states such as waiting in port and anchoring. Combustion mode switching strategy, including the timing and conditions for switching between diesel mode, gas mode and dual-fuel mode. Propulsion power distribution strategy, which is the power distribution scheme between multiple propellers or multi-axis propulsion systems. Propeller pitch change strategy, which is the curve and rate setting of the pitch angle change in the adjustable pitch propeller. Electric propulsion power distribution strategy, which is the distribution ratio of the motor output power in a hybrid or all-electric propulsion system. These control variables together constitute the output structure of the expected decision.

[0047] When constructing the cost function, the multi-objective optimization model uses speed, navigation stability, range mission completion, and energy consumption as comprehensive optimization objectives. Specifically, the cost function includes: a speed deviation cost, which measures the degree of deviation between the actual speed and the speed range required by the mission; an attitude stability cost, which is calculated based on the difference between the predicted pitch and roll angles and the allowable range, reflecting navigation comfort and safety; a mission completion cost, which uses the difference between the estimated arrival time and the planned arrival time as the core indicator; and a unit energy consumption cost, which calculates a comprehensive indicator of fuel or electricity consumption per nautical mile, including a penalty item for emission constraints.

[0048] The cost function is not only a static optimization, but also a dynamic calculation, which receives the following three types of data in real time: real-time navigation parameters (speed, attitude, propeller status, etc.) in the fusion navigation status information; disturbance arrival time window and intensity level parameters in the expected navigation status information; speed and stability demand targets in the expected demand information.

[0049] The dynamic calculation process introduces multiple types of constraint models, including: propulsion and combustion actuator dynamic response models, which describe the time delay and transition characteristics from the actuator receiving the control command to the actual output change; power output upper limit and change rate limit, to prevent power mutations in a short period of time from causing mechanical shock or overload; fuel and power supply capacity boundaries, to ensure that the fuel injection amount and power supply do not exceed the sustainable capacity of the system; structural load and vibration thresholds, to avoid structural fatigue damage caused by severe sea conditions or frequent adjustments; emission regulations, to limit the combustion mode and output power settings according to regional emission control standards (ECA areas, etc.).

[0050] Within the disturbance arrival time window, the system automatically imposes a power change smoothing constraint to limit the rate of change of the target power curve, and introduces an impact cost weight related to the disturbance intensity level. By adding a mechanical shock penalty term caused by the disturbance to the cost function, the transient shock risk of the propulsion system and the engine is reduced.

[0051] Input stage: synchronously receiving fused navigation status information, expected navigation status information and expected demand information from the data processing module and the data prediction module; Model construction stage: combining control variables, cost functions and constraints to form a multi-objective nonlinear optimization model; Solution phase: Use a multi-objective optimization algorithm (such as NSGA-II, particle swarm optimization PSO or improved gradient method) to search for the optimal solution set within the constraints; The optimization solution process through the multi-objective optimization model includes: Selection stage: select the solution with the lowest comprehensive cost as the final expected decision based on the weight setting; Output stage: The optimized expected decision is transmitted to the parameter generation module to enter the next step of the control information generation process.

[0052] Parameter Generation Module: Generates control information, including combustion control parameters, propulsion control parameters, and electrical management parameters, based on expected navigation state information and expected decisions. Within the disturbance arrival time window contained in the expected navigation state information, the control information is feedforward shaped and slope-limited according to the intensity level. When a severe environmental mode is detected, a weight matrix corresponding to the type and level is obtained to perform a weighted update on the control information. The generation control information includes: In the process of generating combustion control parameters, the ambient temperature change prediction in the expected navigation status information is used to correct the combustion mode switching strategy in the expected decision; based on the disturbance prediction in the expected navigation status information and its arrival time window and intensity level, the target output power curve in the expected decision is feedforward adjusted; in the process of generating propulsion control parameters, the wave period prediction and the countercurrent intensity change trend prediction are used to optimize the propulsion power distribution and propeller pitch change strategy in the expected decision; in the process of generating electrical management parameters, the port congestion prediction is used to adjust the electric propulsion power distribution strategy in the expected decision.

[0053] The feedforward shaping and slope limiting processing includes: within a preset advance amount of the disturbance arrival time window, using the disturbance intensity level to determine the allowable variation range and change rate of the injection quantity, rail pressure and air-fuel ratio in the combustion control parameters, applying slope limits to the torque command and propeller pitch command in the propulsion control parameters, and adjusting the control parameter curve through a smooth transition function.

[0054] The weighted update includes: monitoring operating parameters including ambient temperature, sea state level, port congestion and mission type by fusing navigation status information with expected navigation status information; when a single operating parameter or a combination of parameters exceeds the corresponding preset threshold, determining to enter a harsh environment mode, and determining the type and level of the harsh environment mode based on the type and magnitude of the exceeded parameters; obtaining a weight matrix corresponding to the type and level, mapping the matrix elements to each control variable of the combustion control parameters, propulsion control parameters and electrical management parameters, applying a proportional correction coefficient to the combustion control parameters to improve combustion stability or fuel atomization effect, applying a proportional correction coefficient to the propulsion control parameters to increase the power stability margin or limit the propeller pitch change rate, and applying a proportional correction coefficient to the electrical management parameters to optimize the electric propulsion power distribution or adjust the load switching sequence.

[0055] Feedforward shaping: Before a disturbance occurs, the control parameters are actively adjusted according to the predicted information to ensure a smooth transition of the system to a new state.

[0056] Slope limiting: It limits the maximum rate of change of control parameters to prevent the mechanical system from causing shock or instability due to overly rapid adjustment.

[0057] Weight matrix: A two-dimensional coefficient matrix used for proportional correction of multiple control variables, which can be dynamically switched according to the environmental mode.

[0058] In one embodiment, the parameter generation module receives the expected decision from the decision optimization module, which includes: target trajectory; target output power; idle state holding time; combustion mode switching strategy; propulsion power allocation strategy; propeller pitch change strategy; electric propulsion power allocation strategy; At the same time, it receives expected navigation status information from the data prediction module, including: disturbance arrival time window; disturbance intensity level; ambient temperature change prediction; wave period prediction; countercurrent intensity change trend prediction; port congestion prediction; pitch angle and roll angle change trend prediction; surge arrival time prediction; through the above data, the module realizes the generation, shaping and weighted correction of control information.

[0059] Combustion control parameter generation is based on the combustion mode switching strategy in the expected decision-making, combined with the prediction of ambient temperature changes to correct the injection timing, injection amount, rail pressure and air-fuel ratio; before the disturbance arrives at the time window, the combustion control amount is pre-adjusted according to the disturbance intensity level to reduce the risk of combustion instability; when the target segment is within the emission control area (ECA), low emission mode parameters are generated first and matched with the switching instructions of gas mode or mixed combustion mode.

[0060] Propulsion control parameters are generated, combined with wave period prediction and backflow intensity change trend prediction, to correct the propulsion power distribution and propeller pitch change strategy, so that the propeller can maintain stable output in both long-period waves and short-period waves; the propeller pitch command curve is optimized to keep its change rate within a controllable range to avoid mechanical fatigue caused by frequent adjustments.

[0061] Electrical management parameters are generated to dynamically allocate electric propulsion power based on port congestion predictions to avoid frequent motor starts / stops in congested conditions; the electrical load switching sequence is adjusted during long periods of low-load operation to reduce transient current shocks.

[0062] Feedforward Shaping: Within the preset lead time of the disturbance arrival time window, the disturbance intensity level is used to calculate the allowable change amplitude and change rate of various control parameters, and the corresponding curve smoothing function (S-curve or exponential decay function) is selected according to the disturbance type (such as surge, backflow, temperature mutation) to adjust the combustion, propulsion and electrical control signals in advance.

[0063] Slope limit processing: imposes a maximum rate of change limit on the torque command and propeller pitch command to prevent mechanical shock or control loop instability caused by large adjustments in a short time.

[0064] Weighted update mechanism, severe environment mode monitoring, by integrating navigation status information with expected navigation status information, real-time monitoring of operating parameters such as ambient temperature, sea condition level, port congestion, and mission type; if a single or combined parameter exceeds the preset threshold, it is determined to enter the severe environment mode, and its mode type and level are determined based on the type and magnitude of the exceeded parameters.

[0065] The weight matrix application calls the corresponding weight matrix from the preset library according to the type and level of the severe environment mode; the weight matrix elements are mapped to each control variable, and the combustion control, propulsion control and electrical management parameters are corrected proportionally.

[0066] Control execution module: After receiving the control information output by the parameter generation module, it combines the disturbance arrival time window and intensity level contained in the expected navigation status information, as well as the dynamic response characteristics of the propulsion and combustion actuators, to make real-time fine-tuning of the combustion control parameters and propulsion power distribution, and updates the engine control execution instructions based on the fine-tuned control information.

[0067] The real-time fine-tuning includes: Based on the disturbance arrival time window and intensity level, combined with the dynamic response characteristics of the propulsion and combustion actuators, the preset advance amount before the disturbance arrives and the corresponding adjustment amplitude are calculated; in the real-time fine-tuning of the combustion control parameters, the instantaneous change amplitude of the injection quantity, rail pressure and air-fuel ratio are adjusted according to the disturbance intensity level, and a gradual correction is implemented before the disturbance arrives; in the real-time fine-tuning of the propulsion power distribution, the change rate of the torque command and the propeller pitch command is limited according to the dynamic response characteristics of the actuator, and a smooth output is maintained during the disturbance effect. The execution instructions of the engine control are synchronously updated according to the control information after fine-tuning, so as to maintain the optimized state of navigation stability and energy efficiency before, during and after the disturbance arrives.

[0068] In one embodiment, the control execution module fine-tunes the control information in real time during execution to ensure stability and energy efficiency under external disturbances and dynamic changes of the equipment.

[0069] Based on the disturbance arrival time window and intensity level in the expected navigation status information, the preset lead time for intervention is calculated (for example, starting to adjust the combustion and propulsion strategy 15 seconds before the disturbance); the lead time calculation will integrate the dynamic response time of the propulsion actuator and the combustion actuator to ensure that the adjustment action is precisely synchronized with the arrival time of the disturbance.

[0070] When the disturbance intensity level is high, the gradual correction amplitude of the injection quantity and rail pressure is increased, and the air-fuel ratio is appropriately adjusted to prevent flameout or incomplete combustion in the combustion chamber; a smooth transition function is applied to the injection time curve to make parameter changes continuously controllable and reduce engine transient stress.

[0071] The rate of change limit based on the actuator response model is imposed on the torque command and the propeller pitch command to prevent mechanical shock to the propulsion system; the propulsion power curve is kept smooth during the disturbance to avoid frequent fluctuations that cause increased fuel consumption and decreased propulsion efficiency.

[0072] Dynamically prioritize electric propulsion power and non-propulsion electrical loads, for example, prioritizing the power supply to the propulsion motor during large surges or backflows; if transient fluctuations in the power grid occur, the buffer energy module (energy storage unit or flywheel energy storage) is activated to smooth out load impacts.

[0073] The fine-tuned combustion control parameters, propulsion control parameters and electrical management parameters are synchronously sent to the engine control unit (ECU), propulsion control unit (PCU) and power management system (PMS) through the ship control bus (such as CAN bus or industrial Ethernet).

[0074] Timestamp synchronization is performed across multiple control units to ensure coordinated and consistent actions across all actuators. The control execution module compares the deviation between actual operating parameters and target control parameters in real time. If the deviation exceeds the set threshold, a rapid correction process is triggered. During the correction process, navigation stability is prioritized, followed by energy efficiency optimization, and finally secondary objectives such as comfort are considered. If the real-time monitored environment and disturbance characteristics deviate significantly from the expected prediction, the module will request the parameter generation module to quickly recalculate and issue new control instructions within seconds to achieve dynamic adaptation.

[0075] Disturbance lead: refers to the amount of time that control adjustments are made before a disturbance arrives, to ensure that the control effect accurately matches the arrival time of the disturbance.

[0076] Gradual correction: refers to the adjustment of control parameters in a smooth curve to avoid mechanical shock or combustion instability caused by sudden changes.

[0077] Fast recalculation: refers to the ability to recalculate control information and issue instructions in a short period of time when the actual situation deviates significantly from the prediction.

[0078] Through the above design, the control execution module can not only implement advance intervention before the disturbance arrives, but also maintain navigation stability during and after the disturbance occurs, and continuously optimize the control effect through closed-loop feedback, forming a dynamic and adaptive ship power control closed-loop system.

[0079] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A ship engine control system, characterized in that: include: Data processing module: acquires multi-source data of the ship during navigation, performs health monitoring, trust analysis and weighted fusion processing on it, and obtains fused navigation status information; Data prediction module: obtains expected navigation status information and expected demand information in sequence; Decision Optimization Module: Based on the integration of navigation status information, expected navigation status information and expected demand information, a multi-objective optimization model including speed, navigation stability, range mission and energy consumption is established. The power change smoothing constraint and impact cost weight are introduced within the disturbance arrival time window contained in the expected navigation status information to determine the ship's expected decision. Parameter Generation Module: Generates control information, including combustion control parameters, propulsion control parameters, and electrical management parameters, based on expected navigation state information and expected decisions. Within the disturbance arrival time window contained in the expected navigation state information, the control information is feedforward shaped and slope-limited according to the intensity level. When a severe environmental mode is detected, a weight matrix corresponding to the type and level is obtained to perform a weighted update on the control information. Control execution module: After receiving the control information output by the parameter generation module, it combines the disturbance arrival time window and intensity level contained in the expected navigation status information, as well as the dynamic response characteristics of the propulsion and combustion actuators, to make real-time fine-tuning of the combustion control parameters and propulsion power distribution, and updates the engine control execution instructions based on the fine-tuned control information.

2. A ship engine control system according to claim 1, characterized in that: Obtaining the fused navigation status information includes: Perform health monitoring on each source device of data acquired through sensors in multi-source data to generate a corresponding health status; when the health status of a source device is abnormal, trigger virtual compensation to generate alternative data; the health monitoring includes detecting data drift, signal fluctuation amplitude, abnormal delay and data loss; Conduct trustworthy analysis on multi-source data, including calculating the health metrics between similar data based on the health status of different source devices of the same data; verify the association of different types of data through the data association relationship table, calculate the association metrics between different types of data, and calculate the credibility factor of each type of data based on the weighted health metrics and association metrics; obtain the corresponding credibility weight based on the credibility factor, and perform weighted fusion processing on the multi-source data to obtain fused navigation status information.

3. A ship engine control system according to claim 1, characterized in that: The virtual compensation includes: Obtain the health status of the remaining source devices of the same type as the target source device. When the proportion of devices with a healthy health status is higher than a first threshold and the proportion of devices with an abnormal health status is lower than a second threshold, perform fitting estimation based on the data corresponding to the healthy devices to generate alternative data corresponding to the current source device; otherwise, call the estimation model to perform inference based on historical data and relevant data in the data association relationship table. The estimation model is a time series prediction model based on machine learning.

4. A ship engine control system according to claim 1, characterized in that: The sequential acquisition of expected navigation status information and expected demand information includes: performing navigation prediction based on the fused navigation status information, including surge arrival time prediction, ambient temperature change prediction, port congestion prediction, pitch angle and roll angle change trend prediction, wave period prediction and countercurrent intensity change trend prediction, and calculating the disturbance arrival time window and intensity level accordingly, and generating expected navigation status information including the above-mentioned prediction parameters; based on the fused navigation status information and expected navigation status information, combined with historical navigation path data and mission requirement path data, performing speed requirement and stability requirement prediction to generate expected demand information.

5. A ship engine control system according to claim 1, characterized in that: The multi-objective optimization model parameterizes the control variables during the ship operation process into expected decisions including target track, target output power, idle condition maintenance time, combustion mode switching strategy, propulsion power allocation strategy, propeller pitch change strategy and electric propulsion power allocation strategy, and constructs a cost function in the multi-objective optimization model with speed, navigation stability, range mission and energy consumption as comprehensive optimization objectives; the cost function is dynamically calculated based on the real-time navigation parameters in the fused navigation status information, the disturbance arrival time window and intensity level parameters contained in the expected navigation status information, and the expected demand information.

6. A ship engine control system according to claim 5, characterized in that: The dynamic calculation includes: introducing dynamic response models of propulsion and combustion actuators, power output upper limits and change rate limits, fuel and power supply capacity boundaries, structural load and vibration thresholds, and emission regulations limits; and applying power change smoothing constraints within the disturbance arrival time window to limit the change rate of the target power curve. At the same time, an impact cost weight associated with the disturbance intensity level is introduced to reduce the transient mechanical impact of the propulsion system and engine. An optimization solution that minimizes the comprehensive cost function under the above constraints is obtained through a multi-objective optimization solution method, and the optimization solution is output to a parameter generation module.

7. A ship engine control system according to claim 1, characterized in that: The generation control information includes: In the process of generating combustion control parameters, the ambient temperature change prediction in the expected navigation status information is used to correct the combustion mode switching strategy in the expected decision; based on the disturbance prediction in the expected navigation status information and its arrival time window and intensity level, the target output power curve in the expected decision is feedforward adjusted; in the process of generating propulsion control parameters, the wave period prediction and the countercurrent intensity change trend prediction are used to optimize the propulsion power distribution and propeller pitch change strategy in the expected decision; in the process of generating electrical management parameters, the port congestion prediction is used to adjust the electric propulsion power distribution strategy in the expected decision.

8. A ship engine control system according to claim 1, characterized in that: The feedforward shaping and slope limiting processing includes: within a preset advance amount of the disturbance arrival time window, using the disturbance intensity level to determine the allowable variation range and change rate of the injection quantity, rail pressure and air-fuel ratio in the combustion control parameters, applying slope limits to the torque command and propeller pitch command in the propulsion control parameters, and adjusting the control parameter curve through a smooth transition function.

9. A ship engine control system according to claim 1, characterized in that: The weighted update includes: monitoring operating parameters including ambient temperature, sea state level, port congestion and mission type by fusing navigation status information with expected navigation status information; when a single operating parameter or a combination of parameters exceeds the corresponding preset threshold, determining to enter a harsh environment mode, and determining the type and level of the harsh environment mode based on the type and magnitude of the exceeded parameters; obtaining a weight matrix corresponding to the type and level, mapping the matrix elements to each control variable of the combustion control parameters, propulsion control parameters and electrical management parameters, applying a proportional correction coefficient to the combustion control parameters to improve combustion stability or fuel atomization effect, applying a proportional correction coefficient to the propulsion control parameters to increase the power stability margin or limit the propeller pitch change rate, and applying a proportional correction coefficient to the electrical management parameters to optimize the electric propulsion power distribution or adjust the load switching sequence.

10. A ship engine control system according to claim 1, characterized in that: The real-time fine-tuning includes: Based on the disturbance arrival time window and intensity level, combined with the dynamic response characteristics of the propulsion and combustion actuators, the preset advance amount before the disturbance arrives and the corresponding adjustment amplitude are calculated; in the real-time fine-tuning of the combustion control parameters, the instantaneous change amplitude of the injection quantity, rail pressure and air-fuel ratio are adjusted according to the disturbance intensity level, and a gradual correction is implemented before the disturbance arrives; in the real-time fine-tuning of the propulsion power distribution, the change rate of the torque command and the propeller pitch command is limited according to the dynamic response characteristics of the actuator, and a smooth output is maintained during the disturbance effect. The execution instructions of the engine control are synchronously updated according to the control information after fine-tuning, so as to maintain the optimized state of navigation stability and energy efficiency before, during and after the disturbance arrives.

Images