Ship power system energy efficiency optimization control system based on multi-sensor fusion

Abstract

The application relates to the technical field of ship power control, and discloses a ship power system energy efficiency optimization control system based on multi-sensor fusion, which comprises the following modules: a multi-source data acquisition module, which is used for synchronously collecting hull, shafting and air path parameters; a kinematics mapping calculation module, which is used for calculating a propeller transient immersion index; a phase extraction and mode determination module, which is used for extracting a wave phase and an amplitude output mode signal; a fuel mechanical dynamics feedforward module, which is used for generating an oil injection instruction according to the amplitude and the immersion index derivative; an air thermodynamics cooperative control module, which is used for controlling a valve and starting a scavenging blower according to a phase angle; and a long-term energy efficiency optimization module, which is used for finely adjusting a main rotating speed setting value according to an energy efficiency cost function gradient. The technical scheme of extracting the transient immersion index and the fluctuation amplitude to dynamically adjust the rotating speed control dead zone and superimpose a negative fuel feedforward instruction is adopted, so that the technical effects of suppressing abnormal fluctuation of the main engine rotating speed under high sea conditions and reducing unnecessary fuel consumption are achieved.

Description

Technical Field

[0001] This invention relates to the field of marine power control technology, specifically to a marine power system energy efficiency optimization control system based on multi-sensor fusion. Background Technology

[0002] Large ships face complex and ever-changing marine environments during long-distance voyages. The violent rolling and heaving caused by waves lead to constant changes in the water depth of the propeller, resulting in significant fluctuations in engine load. This physical disturbance not only affects the safety of navigation but also directly increases fuel consumption. Furthermore, the hull condition and internal mechanical structure of a ship slowly change over its long service life. To ensure stable power output and excellent fuel economy in harsh sea conditions and throughout its entire lifespan, precise control and adaptive optimization of the power system have become urgent practical needs in the shipping industry.

[0003] Existing marine engine control typically employs independent closed-loop PID control systems and fixed operating chart strategies. This control system can collect real-time engine speed deviations and perform direct internal feedback calculations, thus maintaining relatively stable shafting operation under calm sea conditions. Regarding airflow control, current technology monitors the intake status by setting a fixed lower limit for scavenging pressure, automatically triggering auxiliary operating equipment when a specific pressure threshold is detected. For energy efficiency management, existing solutions generally rely on fixed MAP charts calibrated during factory sea trials for lookup matching. This lookup control method consumes minimal computational resources, has simple and stable program execution logic, and can quickly provide standard speed setting references for different speed commands.

[0004] The aforementioned speed control method relying on internal closed-loop feedback is essentially a reactive adjustment. High sea states cause the propeller to frequently enter and exit the water. The governor only begins to adjust the fuel injection quantity after the actual speed drops or surges significantly. The mechanical actuator itself has a physical action time lag. This results in a slow response of the system to high-frequency wave disturbances, leading to large speed fluctuations and unnecessary fuel consumption. In addition, the exhaust gas turbocharger rotor has a large moment of inertia. It is difficult to synchronously establish scavenging pressure when the load changes abruptly. The logic of passively triggering the air circuit action based on a single pressure limit requires intervention only when the detected value falls below the bottom line. At this time, the combustion chamber is already in an oxygen-deficient state. This not only worsens transient combustion but also easily disrupts the flow stability inside the turbocharger and causes surge. Furthermore, fixed MAP charts cannot reflect the changes in physical boundaries brought about by long-term ship service. Increased drag due to biofouling on the hull and mechanical wear of equipment will change the true energy consumption characteristics of the system. Continuing to rely on fixed charts calibrated many years ago for matching causes the system to lose its adaptive ability to mechanical aging and environmental changes, ultimately resulting in long-term overall energy efficiency decline. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a multi-sensor fusion-based energy efficiency optimization control system for marine power systems. This system solves the problems of existing marine control systems, such as drastic speed fluctuations and abnormal fuel consumption caused by post-event feedback adjustments in harsh sea conditions, combustion degradation and turbocharger surge caused by passive intervention in the gas path, and the inability of fixed operating chart matching to adapt to the overall energy efficiency degradation caused by long-term marine service.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a ship propulsion system energy efficiency optimization control system based on multi-sensor fusion, comprising: The multi-source data acquisition module is used to synchronously acquire and align timestamps to obtain hull kinematic parameters, shaft dynamic parameters, and engine thermodynamic gas path parameters. The kinematic mapping calculation module is used to map the kinematic parameters of the hull to the coordinate system of the bottom propeller relative to the water surface, and calculate the transient immersion index that characterizes the change of the effective work projection area of ​​the propeller. The phase extraction and mode determination module is used to extract the wave encounter phase angle and wave amplitude from the transient immersion index, and output the high sea state cooperative mode signal based on the wave amplitude; The fuel mechanical dynamics feedforward module is used to dynamically expand the control dead zone parameters of the engine control system speed closed loop according to the fluctuation amplitude when receiving the high sea state cooperative mode signal, and generate the total fuel injection command according to the first time derivative of the transient immersion index. The air thermodynamics coordinated control module is used to control the exhaust gas bypass proportional valve to reduce the valve opening according to the wave encounter phase angle when receiving the high sea state coordinated mode signal, and output a lock-up start-on command to the auxiliary scavenging blower. The long-term energy efficiency optimization module is used to calculate the gradient of the energy efficiency cost function composed of specific fuel consumption with respect to the main speed setpoint based on the shaft dynamic parameters and the engine thermodynamic air circuit parameters, and to continuously fine-tune the main speed setpoint according to the descent path of the gradient.

[0007] Preferably, the multi-source data acquisition module is specifically used for: Extract the raw timestamp of the data packet; Use the engine control system master clock pulse as a unified time alignment reference; Extract the current reference timestamp, and retrieve the adjacent sampled data of the hull kinematic parameters, the shaft dynamic parameters, and the engine thermodynamic gas path parameters on both sides of the current reference timestamp in the data buffer area; Linear interpolation is performed based on the current reference timestamp to output an aligned sequence of synchronization state data.

[0008] Preferably, the kinematic mapping calculation module is specifically used for: Extract the instantaneous pitch angle and instantaneous heave displacement from the kinematic parameters of the hull; The transient vertical displacement of the propeller shaft relative to the static equilibrium position of the hull is calculated based on the set static parameters of the ship's geometric dimensions, the instantaneous pitch angle, and the instantaneous heave displacement. The instantaneous absolute immersion depth of the propeller is obtained by adding the transient vertical displacement to the still water stern draft. When the instantaneous absolute immersion depth is between zero and twice the effective physical radius of the propeller, the transient proportion of the effective work projection area is derived as the transient immersion index based on the integral calculation model.

[0009] Preferably, the phase extraction and mode determination module is specifically used for: The transient immersion index is filtered by a high-pass filter to remove the DC bias component and extract the AC disturbance component. The AC disturbance component is input into the built-in phase-locked loop algorithm module to calculate the phase deviation between the AC disturbance component and the reference signal to achieve frequency and phase locking, and output the wave encounter phase angle in real time. The fluctuation amplitude is extracted by using an orthogonal second-order generalized integrator to generate an original signal in phase with the AC disturbance component and an orthogonal signal with a 90-degree lag. The Euclidean norm of the original signal and the orthogonal signal is calculated.

[0010] Preferably, the phase extraction and mode determination module is further used for: Create a sliding time window of a preset length on the timeline; Within the sliding time window, the root mean square (RMS) value of the fluctuation amplitude is calculated by performing root mean square (RMS) statistics on the fluctuation amplitude. When the comparison determines that the root mean square value is greater than the preset sea state threshold parameter, the high sea state cooperative mode signal is output.

[0011] Preferably, the fuel mechanical dynamics feedforward module is specifically used for: The fluctuation amplitude is converted into the speed deviation dead zone increment based on the dead zone adaptive gain constant, so as to adjust the control dead zone parameter; The first time derivative of the transient immersion index is compared with a preset negative threshold value; When the first time derivative of the transient immersion index is less than the preset negative threshold, a negative fuel feedforward compensation is generated. Obtain the basic fuel closed-loop command output by the governor, and superimpose the negative fuel feedforward compensation amount with the basic fuel closed-loop command to generate the total fuel injection command.

[0012] Preferably, the air thermodynamic coordinated control module is specifically used for: Determine whether the wave encounter phase angle falls within the set advance phase angle window; When it is determined that the window has been entered, a turn-off feedforward bias is generated. Obtain the basic gas path valve command output by the gas path controller, and algebraically subtract the basic gas path valve command from the shut-off feedforward offset to control the waste gas bypass proportional valve to reduce the valve opening. Throughout the entire high sea state cooperative mode, the start-up command in a locked state is continuously output to the auxiliary scavenging blower.

[0013] Preferably, the air thermodynamic coordinated control module is further used for: Real-time acquisition of feedback data on the absolute pressure of the scavenging manifold and the turbocharger speed from the engine thermodynamic air path parameters; The absolute pressure of the scavenging manifold is compared with the set maximum allowable scavenging pressure, and the turbocharger speed feedback data is compared with the set maximum allowable speed. When it is determined that the absolute pressure of the scavenging manifold is greater than the maximum permissible scavenging pressure or the turbocharger speed feedback data is greater than the maximum permissible speed, the algebraic subtraction calculation path between the basic air circuit valve command and the shut-off feedforward bias is forcibly blocked, and the terminal execution command of the exhaust bypass proportional valve is forcibly locked to the maximum safe pressure relief opening.

[0014] Preferably, the long-term energy efficiency optimization module is specifically used for: The shaft dynamic parameters and engine thermodynamic gas path parameters are smoothed and filtered on a long-period time scale to extract the average fuel consumption mass flow rate and average shaft power under steady-state conditions. The energy efficiency cost function is obtained by dividing the average fuel consumption mass flow rate by the average shaft power. Under the set macro-optimization period, the difference between the energy efficiency cost function and the historical cost function of the previous calculation period is calculated, and the difference is divided by the change in the main speed setting value to calculate the gradient of the energy efficiency cost function with respect to the main speed setting value.

[0015] Preferably, the long-term energy efficiency optimization module is further used for: The master rotation speed setting value is fine-tuned according to the set optimization learning rate constant and the gradient. When the absolute value of the gradient in multiple consecutive calculation cycles is less than the preset convergence tolerance threshold, the step fine adjustment of the main speed setting value is paused, the current main speed setting value is locked, and the current average shaft power is recorded as the locking reference value. The system monitors the average shaft power. When it is determined that the average shaft power deviates from the locked reference value by more than the set environmental reset tolerance rate, the system automatically releases the locked state of the main speed setting value and reactivates the step fine-tuning logic.

[0016] This invention provides an energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion. It has the following beneficial effects: 1. The present invention adopts a technical solution of extracting transient immersion index and fluctuation amplitude to dynamically adjust the speed control dead zone and superimpose negative fuel feedforward command. This achieves the technical effect of smoothing abnormal fluctuations in main engine speed under high sea state and reducing unnecessary fuel consumption. Compared with the existing technology that simply relies on the engine's internal closed-loop PID for post-feedback adjustment, this invention solves the shortcomings of slow speed regulation response and excessive speed overshoot when facing severe sea state.

[0017] 2. The present invention adopts a technical solution based on the wave encounter phase angle determination result to close the exhaust gas bypass proportional valve in advance and force the auxiliary scavenging blower to start. This achieves the technical effect of preventing sudden pressure drop in the gas circuit system and ensuring sufficient air intake in the combustion chamber. Compared with the existing technology that passively triggers the booster gas circuit action by setting a single scavenging pressure limit, this invention solves the shortcomings of the gas circuit system response lag under sudden propeller load and the tendency to cause booster surge.

[0018] 3. The present invention adopts a technical solution that uses the smoothed average fuel flow and average shaft power to construct a cost function and then finely adjusts the main speed setting value step by step according to the gradient difference path. This achieves the technical effect of enabling the ship's main power system to adaptively follow the lowest energy consumption point. Compared with the existing technology that relies on the factory test to fix the MAP chart to match the speed operation, this invention solves the problem that it cannot cope with the overall energy efficiency decline of the system caused by mechanical aging and environmental changes after long-term service of the ship. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the system structure according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the method flow according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the multi-source data synchronous acquisition logic in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the kinematic derivation logic of the transient immersion index in an embodiment of the present invention. Figure 5 This is a schematic diagram of the wave load feature extraction and mode determination logic in an embodiment of the present invention. Figure 6 This is a schematic diagram of the mechanical dynamics feedforward oil reduction and overshoot prevention control logic of an embodiment of the present invention; Figure 7 This is a schematic diagram of the air thermodynamic dual-track cooperative predictive compensation logic in an embodiment of the present invention. Figure 8 This is a flowchart illustrating the baseline state stripping and long-term energy efficiency optimization logic of an embodiment of the present invention. Figure 9 The diagram shows the system dynamic response curves of a ship entering a wave region from calm water, as described in this embodiment of the invention. (a) is the curve of the transient immersion index and wave encounter phase changing with time, and (b) is the curve of the engine speed and total fuel injection command changing with time. Figure 10 The diagram shows the air thermodynamic compensation and long-term optimization curves of an embodiment of the present invention, where (a) is the curve of the change of scavenging pressure and exhaust bypass proportional valve opening over time, and (b) is the iterative curve of the change of energy efficiency cost function and main speed setpoint over time.

[0020] Among them, 10 is the multi-source data acquisition module; 20 is the kinematic mapping calculation module; 30 is the phase extraction and mode determination module; 40 is the fuel mechanical dynamics feedforward module; 50 is the air thermodynamics coordinated control module; and 60 is the long-term energy efficiency optimization module. Detailed Implementation

[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Reference Figure 1 The present invention provides a ship power system energy efficiency optimization control system based on multi-sensor fusion, which may include: a multi-source data acquisition module 10, a kinematic mapping calculation module 20, a phase extraction and mode determination module 30, a fuel mechanical dynamics feedforward module 40, an air thermodynamics coordinated control module 50, and a long-term energy efficiency optimization module 60.

[0023] The multi-source data acquisition module 10 is configured in the ship's field equipment communication network. The multi-source data acquisition module 10 establishes hardware bus connections with the hull inertial measurement unit, the stern shaft torque meter, and the engine control system. The multi-source data acquisition module 10 synchronously acquires hull kinematic parameters, shafting dynamic parameters, and engine thermodynamic gas path parameters according to a set hardware sampling clock cycle. The multi-source data acquisition module 10 performs timestamp alignment processing on the input data sequence.

[0024] The kinematic mapping calculation module 20 is communicatively connected to the multi-source data acquisition module 10. The kinematic mapping calculation module 20 receives the ship's kinematic parameters output by the multi-source data acquisition module 10. The kinematic mapping calculation module 20 embeds spatial rigid body translation matrix operation logic. Combining the set static parameters of the ship's geometric dimensions, the kinematic mapping calculation module 20 maps the spatial motion attitude at the ship's center of gravity to the coordinate system of the bottom propeller relative to the water surface. The kinematic mapping calculation module 20 calculates the transient immersion index, which characterizes the change in the projected area of ​​the propeller's effective work.

[0025] The phase extraction and mode determination module 30 is connected to the kinematic mapping calculation module 20. The phase extraction and mode determination module 30 receives continuous sequence data of the transient immersion index. The phase extraction and mode determination module 30 calculates the first-order time derivative of the transient immersion index. The phase extraction and mode determination module 30 tracks and separates the transient immersion index sequence through a built-in phase-locked loop algorithm module, obtaining the wave encounter phase angle and the corresponding wave amplitude. Based on the root mean square calculation result of the wave amplitude within a preset time window, the phase extraction and mode determination module 30 determines whether the control system is in the horizontal water mode or the high sea state cooperative mode.

[0026] The fuel mechanical dynamics feedforward module 40 is connected to the phase extraction and mode determination module 30 and the underlying engine control system. When the phase extraction and mode determination module 30 outputs a high sea state cooperative mode signal, the fuel mechanical dynamics feedforward module 40 dynamically expands the control dead zone parameters of the engine control system's speed closed-loop circuit based on the currently extracted fluctuation amplitude. The fuel mechanical dynamics feedforward module 40 generates a negative fuel feedforward compensation amount based on the first time derivative of the transient immersion index. The fuel mechanical dynamics feedforward module 40 superimposes the fuel feedforward compensation amount with the basic closed-loop output result to generate a total injection command, and sends the total injection command to the fuel injection execution channel of the engine control system.

[0027] The air-thermodynamics co-control module 50 establishes communication control connections with the phase extraction and mode determination module 30 and the air path actuators of the engine control system. The air-thermodynamics co-control module 50 acquires real-time feedback data on the absolute pressure of the scavenging manifold and the turbocharger speed transmitted by the multi-source data acquisition module 10. The air-thermodynamics co-control module 50 is equipped with dual-rail output branches. Under high sea state co-control mode, the air-thermodynamics co-control module 50 continuously outputs a lock-up start-up command to the auxiliary scavenging blower based on the long-period average comparison result. The air-thermodynamics co-control module 50 determines whether the current time point is within the set advance phase angle window based on the wave encounter phase angle. When it determines that it has entered the advance phase angle window, the air-thermodynamics co-control module 50 generates a shut-off feedforward bias to control the exhaust bypass proportional valve to reduce the valve opening. The air-thermodynamics co-control module 50 internally contains over-limit interlock protection judgment conditions for absolute scavenging pressure and turbine speed. When the over-limit condition is triggered, the air thermodynamics cooperative control module 50 forcibly blocks the normal algebraic subtraction calculation path, bypasses the output of the basic closed-loop command and the feedforward bias, and forcibly locks the terminal execution command to the maximum safe pressure relief opening.

[0028] The long-term energy efficiency optimization module 60 is connected to both the multi-source data acquisition module 10 and the engine control system's speed setting channel. The long-term energy efficiency optimization module 60 accumulates the operating data acquired by the multi-source data acquisition module 10 over a long-term time scale, calculating the average fuel consumption mass flow rate and average shaft power. The long-term energy efficiency optimization module 60 calculates the gradient of the energy efficiency cost function, composed of specific fuel consumption, with respect to the speed setting value. The long-term energy efficiency optimization module 60 continuously fine-tunes the main speed setting value of the engine control system according to the gradient descent path.

[0029] Reference Figure 2 This invention provides a method for optimizing the energy efficiency control of a ship's propulsion system based on multi-sensor fusion, comprising the following steps: S100, high-frequency synchronous acquisition of multi-source data: synchronously acquires ship motion attitude data, shaft dynamics data and thermodynamic gas path data at a set period through the field hardware communication bus, and performs timestamp alignment operation. S200, kinematic derivation of transient immersion index: The ship's center of gravity motion data is mapped to the propeller coordinate system by rigid body coordinate system translation calculation logic, and the transient immersion index, which represents the effective work projection area of ​​the propeller, is calculated. S300, Wave Load Feature Extraction and Mode Determination: Extract wave encounter phase angle and wave amplitude by differentiating the transient immersion index and performing phase-locked loop tracking. Determine the operating mode of the control system based on the wave amplitude statistics. S400, Mechatronics Feedforward Fuel Reduction and Overshoot Prevention Control: Adaptively adjusts the engine speed closed-loop dead zone parameters under high sea state cooperative mode, and issues feedforward reduction commands to the bottom fuel injection channel based on the transient immersion index change rate. S500, air thermodynamic dual-track coordinated predictive compensation: long-term locking of the auxiliary scavenging blower state under high sea state coordinated mode, and fine adjustment of the exhaust bypass proportional valve for air path pre-pressurization within the set wave phase advance window, while performing physical safety boundary interlock protection. S600, baseline state stripping and long-term energy efficiency optimization: smooths system operating parameters over a long-term time scale and adjusts the engine main speed setpoint along the gradient descent direction of the energy efficiency cost function.

[0030] Subsequent chapters will provide a detailed explanation of each of the above control steps, combining specific algorithm derivations and variable passing logic.

[0031] Reference Figure 3 In the energy efficiency optimization control method for ship propulsion systems based on multi-sensor fusion, when performing the high-frequency synchronous acquisition step S100 of multi-source data, the multi-source data acquisition module 10 specifically performs the following steps: S101, Configure the hardware communication network interface. In this embodiment, the multi-source data acquisition module 10 establishes a data interaction connection with the underlying sensor nodes through a field communication bus. Specifically, this module connects to the inertial measurement unit located at the ship's center of gravity via a serial communication interface to acquire the ship's spatial motion attitude data; connects to the stern shaft torque meter via an analog-to-analog conversion module to acquire shafting mechanical load data; and connects to the engine control system via an industrial Ethernet protocol to acquire the status data of the air circuit and fuel actuators. For the specific selection of communication cables and the configuration of the underlying bus protocol, those skilled in the art can make conventional settings according to the actual environment of the ship. The physical implementation of the communication interface is a well-known technology in the field and will not be described in detail here.

[0032] S102, Set the discrete high-frequency sampling clock. The multi-source data acquisition module 10 is internally configured with a fixed-period timer interrupt trigger to establish a fixed discrete sampling period. As a preferred approach, to match the dynamic frequency changes of wave loads and satisfy the sampling theorem, while avoiding aliasing of high-frequency disturbance signals, this discrete sampling period... The value range is typically set to 10 milliseconds to 50 milliseconds. This applies to each discrete sampling period. Upon arrival, the system issues a synchronization data request command to obtain the real-time state variables distributed across each subsystem. The input vector obtained here is based on the current time. The definition mainly covers three data dimensions: hull kinematic parameters derived from the inertial measurement unit, including instantaneous pitch angle. With instantaneous heave displacement Shaft dynamic parameters derived from the stern shaft sensor, including instantaneous torque. Compared with actual crankshaft speed ; derived from the thermodynamic and airflow parameters of the engine control system, including the absolute pressure of the scavenging manifold. Turbocharger speed and feedback on the current opening degree of the exhaust bypass proportional valve .

[0033] S103, perform heterogeneous data timestamp alignment processing. From an engineering implementation perspective, due to the inherent differences in data update rates and transmission delays among nodes in the ship's fieldbus network, directly introducing unaligned discrete parameters can easily lead to timing misalignments in subsequent spatial kinematic mapping and phase extraction stages. Therefore, when the multi-source data acquisition module 10 receives data packets uploaded by various sensors, it synchronously extracts the original timestamp of the data generation and establishes a circular data buffer with first-in-first-out characteristics to store data sequences with the original timestamps. Regarding the selection of the alignment reference, the system selects the engine control system's master clock pulse as the unified time alignment reference. In the specific calculation process, the current reference timestamp is extracted. It then retrieves the nearest sampled data for each state variable on either side of the reference timestamp from the buffer. For any original discrete sequence of state variables, a reference timestamp is set. satisfy Then, the aligned state variable values ​​are calculated according to the following formula. : ; in, The most recent original sampling time before the reference timestamp. The most recent original sampling time after the baseline timestamp. for The state variable value at that moment. for The state variable value at any given time.

[0034] Furthermore, to ensure the integrity of the algorithm logic and avoid interpolation distortion caused by packet loss due to communication congestion, the multi-source data acquisition module 10 internally presets a maximum allowable delay threshold. Before performing the above interpolation, if the time interval is determined... Greater than the maximum allowable delay threshold If the original data is deemed invalid, the system discards the current linear interpolation calculation and directly outputs the valid historical value from the previous period. Under normal communication conditions, this module calculates the original data for each dimension sequentially according to the above method, completes the numerical mapping of each state variable on the same reference time axis, and outputs an aligned sequence of synchronized state data.

[0035] Reference Figure 4 In the process of performing the transient immersion index kinematics derivation step S200, the kinematic mapping calculation module 20 of the ship power system energy efficiency optimization control method based on multi-sensor fusion specifically performs the following steps: S201, Establish the translation matrix of the spatial rigid body coordinate system. From a physical perspective, the ship's inertial measurement unit is typically installed at the geometric center of gravity of the hull, while the propeller is located at the stern. When the ship undergoes spatial attitude changes in waves, the pitching motion of the hull amplifies the vertical displacement of the stern. To eliminate this spatial positional difference, the kinematic mapping calculation module 20 receives the time-aligned instantaneous pitch angle output by the multi-source data acquisition module 10. With instantaneous heave displacement In this embodiment, the coordinate system rules are set as follows: heave displacement is positive upwards along the vertical line, and pitch angle is positive in the direction of bow pitch. Based on the set static geometric parameters of the ship, this module extracts the horizontal projection distance from the physical geometric center of the inertial measurement unit installation to the longitudinal center plane of the propeller. As a preferred method, this horizontal projection distance Fixed numerical values ​​can be obtained by reading the ship's static factory construction drawings. The kinematic mapping calculation module 20 calculates the transient vertical displacement of the propeller shaft center relative to the ship's static equilibrium position using the following formula. : ; in, This represents the relative vertical displacement after spatial mapping. The dynamic change of this value reflects the additional undulation at the stern caused by the rigid body motion of the hull.

[0036] S202 calculates the instantaneous absolute submersion depth of the propeller. Sea state and wave disturbances directly determine the actual underwater depth of the propeller, and this absolute depth is a key physical basis for assessing transient changes in hydrodynamic loads. To facilitate the subsequent geometric calculation of the arcuate area, the submersion depth referred to in this paper is uniformly defined as the lowest point of the propeller sweep circle. The kinematic mapping calculation module 20 obtains the still water stern draft of the ship under the current loading conditions. In practice, the system reads the conventional stern draft value from the ship's onboard computer and subtracts the fixed structural height difference between the lowest point of the propeller and the keel, thus calculating the reference water depth from the sea level to the lowest edge of the propeller in still water. Combining the relative displacement calculation results obtained in the preceding steps, the kinematic mapping calculation module 20 calculates the instantaneous absolute immersion depth of the propeller according to the following formula. : ; in, It represents the absolute vertical height of the lowest edge of the propeller (the bottom of the sweep circle) from the actual sea level at the current time point.

[0037] S203 performs the piecewise mapping derivation of the transient immersion index. During high sea state navigation, the propeller frequently enters a partially submerged state. Due to the difference in fluid medium density, propeller submersion causes a sudden change in the effective work projection area in actual contact with the water, resulting in a decrease in load torque and causing dynamic misalignment at the front end of the engine shaft system. To quantify this physical phenomenon, the kinematic mapping calculation module 20 introduces the transient immersion index. , used to characterize the transient proportion of the projected area of ​​the effective hydrodynamic work done by the propeller.

[0038] The kinematic mapping calculation module 20 internally presets the effective physical radius parameter of the propeller. This module, based on the principle of calculating the area of ​​a circular segment, constructs a three-segment nonlinear mapping function for different underwater depth states of the propeller. Specifically, when the propeller is determined to be in a fully submerged state, i.e., the determination condition is met... At that time, the kinematic mapping calculation module 20 outputs the transient immersion index at its maximum value: ; When the propeller is determined to be partially above water, i.e., the determination condition is met. At that time, the kinematic mapping calculation module 20 derives the transient proportion of the projected area based on the following integral calculation model: ; When the propeller completely leaves the water surface, the judgment condition is met. At that time, the kinematic mapping calculation module 20 outputs the transient immersion index at its minimum value: ; Using the aforementioned nonlinear mapping model, the kinematic mapping calculation module 20 transforms the multi-source heterogeneous ship spatial motion attitude parameters into a dimensionless normalized exponential sequence. This transient immersion index... It is then continuously transmitted to subsequent control modules, providing prior features for the entire power system to perform cross-domain decoupling and feedforward command allocation.

[0039] Reference Figure 5 In the energy efficiency optimization control method for ship power systems based on multi-sensor fusion, when performing wave load feature extraction and mode determination step S300, the phase extraction and mode determination module 30 specifically performs the following steps: S301, perform the first-order differential calculation of the transient immersion index. In the ship navigation environment, the propeller's entry or exit from the water directly triggers a sudden change in hydrodynamic load. The system needs to quantify this transient rate of change to detect load trends in advance. Considering that the field sensor signals inevitably contain high-frequency white noise caused by mechanical vibration, direct differentiation would lead to spike distortion in the calculation results. Therefore, in this embodiment, the phase extraction and modal determination module 30 receives the discrete transient immersion index sequence output from the previous kinematic derivation. Then, the sequence is preprocessed for noise reduction using a built-in first-order low-pass smoothing filter. Subsequently, the discrete sampling period is set based on the synchronous acquisition of multi-source data. This module calculates the time derivative of the transient immersion index using a first-order backward difference algorithm. The specific calculation formula is as follows: ; in, This is the transient immersion index after low-pass filtering from the previous sampling period. The physical meaning of this derivative value lies in characterizing the rate of change of the effective area of ​​the propeller. When... This indicates that the propeller is in the deep immersion stage, and the fluid load is increasing; when This indicates that the propeller is in the water exit stage, and the fluid load will face the risk of a sudden drop.

[0040] S302, based on a phase-locked loop (PLL), performs feature separation and phase tracking. To extract prior features with periodic patterns from complex wave disturbances, the phase extraction and mode determination module 30 integrates a PLL algorithm unit. In actual operation, the original transient immersion index contains a DC bias component reflecting the current average draft of the ship and an AC disturbance component reflecting wave undulations. The phase extraction and mode determination module 30 removes the DC bias from the sequence data using a high-pass filter, extracting the zero-mean AC disturbance component. This AC disturbance component is then input into the PLL unit, where an internal phase detector calculates the phase deviation between the input signal and the reference signal generated by the voltage-controlled oscillator (VCO), and after smoothing by a loop filter, frequency and phase locking is achieved. In the locked state, the PLL unit outputs the wave encounter phase angle in real time. Its numerical range is limited to the interval [0, 2π].

[0041] Meanwhile, to obtain the instantaneous fluctuation intensity of the wave cycle, the phase extraction and mode determination module 30 uses an orthogonal second-order generalized integrator configured at the front end of the phase-locked loop to generate an original signal in phase with the AC disturbance component. and its orthogonal signal with a 90-degree lag Based on this orthogonal component, the system extracts the fluctuation amplitude within the current wave cycle by calculating the Euclidean norm. The calculation formula is as follows: ; For the specific underlying circuitry or discretization equations of the phase-locked loop algorithm and the orthogonal second-order generalized integrator, those skilled in the art can use well-known techniques to implement them, and will not elaborate further here.

[0042] S303 performs sliding time window statistics and mode switching judgment. The control system needs to determine whether to intervene in the underlying thermodynamic intervention based on the overall severity of the sea state. The phase extraction and mode determination module 30 establishes a time axis with a length of... The sliding time window. To cover the complete wave encounter cycle and avoid short-term sudden noise interference, the length of this sliding time window is... This module is typically set to include 3 to 5 typical wave cycles. It adjusts the fluctuation amplitude within a sliding time window. Perform root mean square statistics and calculate the root mean square value. : ; in, The total number of discrete sampling points within the sliding time window, satisfying .

[0043] The phase extraction and mode determination module 30 has internally preset threshold parameters characterizing the intensity of sea state disturbances. As a preferred approach, this threshold parameter The value can be obtained through offline calibration. It is typically calculated and mapped based on the maximum allowable torque fluctuation rate of the ship's propeller. In this embodiment, a reasonable value range is set between 0.10 and 0.25 (indicating that the root mean square fluctuation of the effective working area reaches 10% to 25%). The system continuously compares the calculated root mean square values. And this threshold parameter. When determining At this point, it indicates that the propeller depth fluctuations caused by the current sea state have reached the threshold that could trigger a power system misalignment. The state machine outputs a trigger command, switching the control system from the conventional level water mode to the high sea state cooperative mode. To ensure the integrity of the algorithm and avoid frequent oscillations and jumps in the state machine under critical sea states, a hysteresis mechanism is introduced into the comparison logic at this point. After switching to the high sea state cooperative mode, a hysteresis mechanism is only introduced when the root mean square value decreases and remains below the lower limit of the hysteresis. hour( (After setting the hysteresis dead zone width), the state machine recovers and outputs the horizontal mode signal. This mode signal will be passed as a global identifier to subsequent mechanical dynamics and aero-thermodynamics control channels.

[0044] Reference Figure 6 In the energy efficiency optimization control method for ship power systems based on multi-sensor fusion, when executing the mechanical dynamics feedforward fuel reduction and overshoot prevention control step S400, the fuel mechanical dynamics feedforward module 40 specifically performs the following steps: S401 executes adaptive dead-zone adjustment of engine speed closed loop based on fluctuation amplitude. When a ship encounters high sea state conditions, the conventional engine speed control system will undergo high-frequency passive adjustment due to transient speed fluctuations caused by waves. Such frequent actions are difficult to counteract external hydrodynamic disturbances and instead increase fuel consumption and mechanical wear of the actuators. To shield against this high-frequency integral accumulation, the fuel-mechanical dynamics feedforward module 40 activates dead-zone expansion logic after receiving the high sea state cooperative mode signal from the phase extraction and mode determination module 30. This module reads the fluctuation amplitude of the current wave cycle. The error dead zone range of the engine base speed closed loop is dynamically adjusted according to the following linear mapping equation. : ; in, This is the inherent base speed dead zone of the engine, and its value is usually determined by the engine manufacturer based on bench tests. This is the dead-zone adaptive gain constant, whose physical dimensions are consistent with the unit of rotational speed (e.g., revolutions per minute), used to convert the dimensionless immersion fluctuation amplitude into the speed deviation dead-zone increment. As a preferred method, The value range can be set to 10% to 20% of the maximum allowable speed fluctuation. In this embodiment, through the above-mentioned adjustment mechanism, when the wave disturbance intensifies, the non-response range of the speed controller is widened synchronously, thereby effectively blocking the passive response of the underlying PID controller to uncontrollable high-frequency wave loads.

[0045] S402 executes the trigger threshold determination and negative feedforward compensation amount generation. In addition to suppressing conventional fluctuations, the control system needs to implement predictive intervention for large gradient load changes. When the propeller accelerates out of the water, the mechanical load decreases rapidly. If conventional negative feedback logic is followed, the engine will only begin to reduce fuel injection after a significant overshoot at the actual speed, posing a risk of speed runaway. Therefore, under the premise of the aforementioned high sea state cooperative mode, the fuel-mechanical dynamics feedforward module 40 monitors the first-order time derivative of the transient immersion index in real time. This module has a preset critical value for the water output rate. This threshold is typically set based on actual ship propeller cavitation critical speed and hydrodynamic characteristic tests. The system will use the current... Compare with the negative critical value. When the condition is satisfied... Upon confirmation of a sudden load drop, the module immediately triggers the feedforward fuel reduction logic and calculates the fuel feedforward compensation amount. : ; When the judgment condition is not met (i.e.) When this occurs, the fuel feedforward compensation is forcibly set to zero, i.e. .in, This is the feedforward compensation coefficient, its dimension being the product of the fuel supply command unit and time (e.g., fuel supply percentage per second). Due to the triggering conditions... The calculated value is negative. This is a negative compensation command. Feedforward compensation coefficient. The selection of compensation parameters needs to be matched with the specific engine's torque response characteristics to ensure that the compensation magnitude accurately offsets the expected reduction in load torque. To avoid mechanical shock caused by sudden changes in compensation, in... During the zeroing-out phase, the module is equipped with a first-order inertial element to achieve smooth instruction decay.

[0046] S403 executes the superposition of the master injection command and sends it to the underlying communication layer. After the feedforward command is generated, it needs to be integrated with the original closed-loop control logic to drive the underlying hardware. The fuel mechanical dynamics feedforward module 40 reads the basic closed-loop output result calculated by the PID governor from the underlying engine control system. The module will use the negative fuel feedforward compensation amount calculated beforehand. The results of this basic closed-loop operation are directly algebraically superimposed at the numerical level to generate the final total fuel injection command executed at the underlying level. : ; To ensure the basic operational safety of the ship's main propulsion system and prevent engine shutdown due to excessive feedforward reduction, the system sets a lower limit clamping logic at the superimposed output node. The fuel mechanical dynamics feedforward module 40 will... The fuel quantity is compared with the idle fuel quantity limit required to maintain the engine's minimum stable speed. If the calculated value is lower than this limit, the system is forced to clamp to the idle fuel quantity limit. After the verification is completed, the fuel mechanical dynamics feedforward module 40 transmits the total fuel injection command via the field communication bus. The fuel injection execution channel is sent to the engine control system. The hardware interaction process of the lower-level controller receiving commands and driving the specific actions of the electronic fuel injection common rail unit can be implemented by those skilled in the art with reference to conventional electronic fuel injection diesel engine control protocols. The specific action control is well-known technology in this field and will not be elaborated upon here.

[0047] Reference Figure 7 In the energy efficiency optimization control method for ship propulsion systems based on multi-sensor fusion, when executing the air-thermodynamic dual-track cooperative predictive compensation step S500, the air-thermodynamic cooperative control module 50 specifically performs the following steps: S501 executes low-frequency baseline bottoming-out lock control. While the aforementioned mechanodynamic feedforward fuel reduction operation can suppress speed overshoot, it causes a decrease in engine exhaust energy, leading to a drop in turbocharger speed and insufficient scavenging pressure. To address this thermodynamic hysteresis, the air-thermodynamic coordinated control module 50 activates the low-frequency baseline control branch in the dual-rail braking path compensation strategy. This module calculates the system's past... Average scavenging pressure within the sliding time window Compare this average value with the baseline scavenging safety limit required at the current average speed. A comparison is performed. In this embodiment, the basic scavenging safety limit... It can be obtained by interpolation from a two-dimensional graph of intake pressure calibration under preset steady-state conditions within the engine control system. When the judgment condition is met... At that time, the air thermodynamics coordinated control module 50 outputs a lock start signal to the auxiliary scavenging blower. As a preferred approach, to avoid electrical fatigue damage caused by frequent motor starts and stops under critical conditions, a long-term locking mechanism is embedded in the control logic. The start signal maintains a constant output state throughout the entire high sea state cooperative mode, refusing to follow the phase of the underlying high-frequency waves. The change occurs only when the control system exits the high sea state cooperative mode, or when the mean scavenging pressure rises and exceeds the upper limit setting of the hysteresis range (e.g., set to...). ,in When the pressure hysteresis constant is set, the locking signal will be released and set to zero, and the auxiliary scavenging blower will stop operating.

[0048] S502 performs high-frequency predictive airflow alignment based on an advance phase angle window. Besides low-frequency support, the system requires core aerodynamic intervention to address transient load impacts from high-frequency waves. From the perspective of engine thermodynamic response principles, when the exhaust gas bypass proportional valve is closed, less exhaust gas flows through the bypass pipe, and more high-temperature, high-pressure exhaust gas is forced into the exhaust turbine, thereby accelerating the coaxial compressor impeller and increasing the absolute pressure in the intake manifold. The aerodynamic thermodynamic co-control module 50 relies on this exhaust gas bypass proportional valve for high-frequency compensation. This module internally presets an advance phase angle window. The specific value of this phase angle window is calculated based on the ratio of the inherent hysteresis time of engine airflow pressure buildup to the current wave encounter period, ensuring that sufficient scavenging back pressure to resist sudden loads is established before the propeller is fully submerged. When the current wave encounter phase angle is detected... If it falls within this interval, then it satisfies This indicates that the propeller is about to fully re-enter the water, and the engine load demand is about to increase. At this time, the system immediately calculates the shut-off feedforward offset of the exhaust bypass proportional valve. : ; in, This is the bias gain constant, whose physical dimension is the ratio of valve opening percentage to radians. When the real-time phase does not fall within the aforementioned advance phase angle window, this turn-off feedforward bias... Forced to zero. To avoid valve mechanical shock and air circuit pressure oscillation caused by sudden changes in bias, in During the zeroing-out phase, a first-order low-pass smoothing circuit is connected in series within the module to achieve a gradual decay of the feedforward bias. Subsequently, the air thermodynamics coordinated control module 50 will use the basic valve opening command generated by the conventional gas path PID controller. The algebraic subtraction of the aforementioned feedforward bias is used to generate the final execution command sent to the waste gas bypass proportional valve. : ; To prevent the calculation results from exceeding the operational boundaries of the physical actuator, the system adds opening lower limit clamping logic before issuing the results. The control command is compared with the valve's minimum safe opening limit (e.g., 0% or a reserved small opening protection value), and the larger of the two values ​​is taken as the actual output command. This control action forces the valve to reduce its opening before the actual physical load arrives, reducing exhaust gas loss and thus forcibly increasing the turbocharger's work capacity to establish the necessary back pressure in the air path. For the opening servo drive mechanism of the bottom-level exhaust bypass proportional valve, those skilled in the art can use a conventional electro-pneumatic converter positioner, which is well-known in the field and will not be elaborated upon here.

[0049] S503 implements a physical absolute boundary over-limit strong-cut interlock protection. If the predictive air pressure boosting operation is affected by abnormal sea conditions, it may cause the intake manifold pressure to exceed the limit, leading to mechanical failure risks such as engine cylinder pressure exceeding the limit. To ensure the physical safety of the equipment, the air-thermodynamics coordinated control module 50 has an absolute safety interlock boundary logic embedded in its underlying layer. This module acquires the absolute scavenging pressure feedback in real time during each control cycle. With turbocharger speed The system will compare the acquired data with the set maximum allowable scavenging pressure. and the maximum allowable speed of the turbine A hard limit comparison is performed. As a preferred method, the aforementioned safety red line threshold... and Static configuration is performed by reading the mechanical alarm limit parameter table in the engine's factory bench test manual. If it is determined that... or In cases of exceeding operating limits, the system executes a coverage intervention operation. The air thermodynamics co-control module 50 immediately and forcibly blocks the conventional algebraic subtraction calculation path, bypassing the basic closed-loop command. With feedforward bias The system outputs [the energy]. Simultaneously, to quickly release excess exhaust energy, the system directly executes commands from the terminal. The maximum safe pressure relief opening is overwritten and forcibly locked (usually set to 100% full opening). When the pressure and speed drop to the lower limit of the safe hysteresis range, the control system must undergo a set delay confirmation period before the feedforward control logic and conventional PID control logic can be allowed to re-engage. This hard logic mechanism serves as a low-level protection, preventing the risk of system damage caused by prediction errors or abnormal wave packets.

[0050] Reference Figure 8 In the energy efficiency optimization control method for ship power systems based on multi-sensor fusion, when performing the baseline state stripping and long-term energy efficiency optimization step S600, the long-term energy efficiency optimization module 60 specifically performs the following steps: S601 performs transient noise stripping and steady-state cost function construction. When a ship operates in complex sea conditions, various high-frequency feedforward and compensation actions implemented by the front-end control system cause instantaneous fluctuations in engine output shaft power and fuel consumption data. Directly using this transient data to evaluate system economy can easily lead to oscillations in optimization direction and misjudgments. Therefore, the long-term energy efficiency optimization module 60 performs smoothing and filtering processing on the underlying high-frequency data over a long-period timescale to strip away transient physical noise and extract steady-state benchmark features reflecting the system's long-term economic performance. This module reads continuously acquired instantaneous shaft power... With instantaneous fuel mass flow rate It also introduces a sliding time window that covers multiple low-frequency disturbance cycles. As a preferred method, to effectively filter out wave interference, the length of this time window is typically set to 5 to 10 times the typical wave encounter period. The system calculates the steady-state average axis power within this time window using an integral averaging algorithm. With steady-state average fuel flow : ; ; The physical dimension of shaft power is kilowatt (kW), and the physical dimension of fuel mass flow rate is kilograms per hour (kg / h). Based on this data stripping and smoothing process, the long-term energy efficiency optimization module 60 establishes a system energy efficiency cost function with specific fuel consumption as the core objective. Its mathematical expression is defined as: ; In the formula, the constant 1000 is used to convert the unit to grams per kilowatt-hour (g / kWh), which is commonly used in the engineering field. Through this cost function, the control system can quantify the overall energy conversion efficiency under the current macroscopic operating state. For the specific measurement and acquisition methods of the underlying instantaneous torque, speed, and fuel flow signals, those skilled in the art can use conventional shaft torque meters and Coriolis mass flow meters. The data acquisition technology is well-known in the field and will not be elaborated further here.

[0051] S602 performs setpoint fine-tuning and convergence based on gradient descent. After establishing the steady-state cost function, the control system performs closed-loop optimization of the main engine's basic control commands according to the evolution trend of this function, aiming to find and lock the economic operating boundary under the current sea state and hull resistance conditions. The long-term energy efficiency optimization module 60 uses the engine main speed setpoint... As an optimization variable, due to the significant mechanical and thermodynamic inertia of the ship's propulsion system and the fact that the cost function calculation depends on a long sliding time window, the long-term energy efficiency optimization module 60 performs iterations according to a macroscopic optimization cycle independent of the underlying high-frequency control cycle. Specifically, after each new speed setpoint is issued by the system, a dwell timer is started, which is set to be greater than the system setpoint time and the sliding time window. The sum is used to ensure that the underlying electromechanical equipment fully enters the new steady-state response range. After the residence period ends, due to the cost function With speed set value This involves a nonlinear hydrodynamic and thermodynamic coupling mechanism, and the module employs a difference method based on numerical perturbation to dynamically estimate the rate of change of the target gradient. The discrete macroscopic optimization step size... Below, the partial derivative of the cost function with respect to the main rotational speed setpoint (i.e., the optimization gradient) The calculation is as follows: ; To prevent program logic exceptions where the denominator is zero during calculation, the system has a small quantity judgment mechanism before performing the difference operation. When the absolute value of the change in the speed setpoint is detected... Less than the set minimum detection step size At this time, the system will actively add a small perturbation to the current speed setpoint, forcing the state to shift effectively. Based on the calculated gradient direction, the long-term energy efficiency optimization module 60 fine-tunes the base speed according to the gradient descent principle, and its setpoint update rule is as follows: ; in, The set learning rate constant is used to control the step size of speed fine-tuning. Its value needs to be matched with the response inertia of the actual ship speed control loop to avoid system instability caused by excessive single adjustment. To ensure the basic requirements of ship navigation missions, the long-term energy efficiency optimization module 60 is internally configured with boundary constraint logic. The system will update the setpoint. Rotation speed range allowed by upper-level navigation missions If the value exceeds the allowable range, the setting will be clamped to the corresponding boundary limit.

[0052] The optimization loop includes a convergence check mechanism, which determines the convergence of gradients within multiple consecutive calculation cycles. Less than the preset convergence tolerance threshold At this point, the system determines that it has reached the local economic operating point under the current operating conditions. The long-term energy efficiency optimization module 60 then pauses the step-by-step fine-tuning of the setpoint and locks the current point. It continuously outputs power to the underlying speed control system. Simultaneously, to adapt to dynamic changes in the external environment, the module's backend continuously monitors the steady-state average shaft power. The fluctuation situation. In this embodiment, if it is determined... If the deviation from the recorded value at the lockout time exceeds the set environmental reset tolerance rate (e.g., 5%), indicating a substantial change in external wind, waves, or resistance, the system will automatically release the setpoint lockout and reactivate the aforementioned gradient optimization process. This mechanism enables closed-loop control of the system, transitioning from passive disturbance rejection to long-term adaptive energy efficiency improvement.

[0053] To further clarify the collaborative working process of the technical solution described in this invention, a specific working scenario example will be used for illustration below.

[0054] The vessel is initially set to operate in calm water. Subsequently, the vessel enters a wave-turbulent area, and the sea state level rises.

[0055] Reference Figure 9 In the initial stage, the ship is in calm water. Figure 9 (a) The transient immersion index remains at a fixed maximum value, indicating that the propeller is fully submerged. When the ship enters the wave zone, the pitching and heaving of the hull cause changes in the propeller's depth relative to the water surface. The kinematic mapping calculation module calculates and outputs the transient immersion index sequence of the wave by receiving sensor data. The phase extraction and mode determination module performs differential processing on this sequence and uses a phase-locked loop to extract the wave encounter phase angle. The system calculates the root mean square of the wave amplitude within a sliding time window. When the root mean square value exceeds a set boundary threshold, the system switches from the level water mode to the high sea state cooperative mode.

[0056] Under high sea state co-mode conditions, the propeller may experience water spillage during certain wave cycles, leading to a decrease in engine mechanical load. For example... Figure 9 As shown in (b), the fuel mechanical dynamics feedforward module is triggered when the first time derivative of the transient immersion index reaches a set negative threshold. This module calculates the negative fuel feedforward compensation and superimposes it with the basic closed-loop command output by the PID governor. The curve correspondence shows that the total injection command begins to decrease before the actual engine speed increases significantly. This action counteracts the excess torque caused by the reduced load, preventing significant speed overshoot.

[0057] Reference Figure 10 When the propeller transitions from an outflow to a reflow state, the hydrodynamic load will increase. For example... Figure 10As shown in (a), the air thermodynamics co-control module monitors the wave encounter phase angle in real time. When the phase angle enters the set advance phase angle window, the system determines that the load is about to increase and then generates a shut-off feedforward bias. This bias is subtracted from the basic closed-loop command, controlling the exhaust bypass proportional valve to reduce its opening. The bypass pipe throttling allows more exhaust gas to enter the turbine, driving the compressor to accelerate. The curve shows that the absolute pressure of the scavenging manifold has already shown an upward trend before the actual sudden increase in mechanical load, completing the air circuit pre-compression operation. During this process, the system continuously sends a lock-on start command to the auxiliary scavenging blower to maintain the reference pressure of the intake system. If the scavenging pressure touches the safety limit, the system blocks the above operation and sets the valve to the maximum opening for pressure relief protection.

[0058] While the underlying system performs high-frequency wave disturbance rejection control, the long-term energy efficiency optimization module performs iterations on the macroscopic time axis. For example... Figure 10 As shown in (b), the system introduces a long sliding time window to perform integral averaging on the collected instantaneous shaft power and instantaneous fuel flow, filtering out high-frequency physical noise. The system uses the processed steady-state data to calculate the energy efficiency cost function characterizing specific fuel consumption. After each set dwell waiting period, the system uses the differential method to calculate the gradient of the cost function with respect to the main engine speed setpoint. The long-term energy efficiency optimization module fine-tunes the engine main engine speed setpoint along the descending direction of the gradient, according to a set step size. After multiple cycles of calculation and distribution, the calculated value of the cost function shows a decreasing trend. When the absolute value of the continuously calculated gradient is less than the convergence tolerance threshold, the system stops adjusting and locks the current speed setpoint. This setpoint is the economic operating point under the current hull resistance and wave environment.

[0059] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A ship propulsion system energy efficiency optimization control system based on multi-sensor fusion, characterized in that, include: The multi-source data acquisition module is used to synchronously acquire and align timestamps to obtain hull kinematic parameters, shaft dynamic parameters, and engine thermodynamic gas path parameters. The kinematic mapping calculation module is used to map the kinematic parameters of the hull to the coordinate system of the bottom propeller relative to the water surface, and calculate the transient immersion index that characterizes the change of the effective work projection area of ​​the propeller. The phase extraction and mode determination module is used to extract the wave encounter phase angle and wave amplitude from the transient immersion index, and output the high sea state cooperative mode signal based on the wave amplitude; The fuel mechanical dynamics feedforward module is used to dynamically expand the control dead zone parameters of the engine control system speed closed loop according to the fluctuation amplitude when receiving the high sea state cooperative mode signal, and generate the total fuel injection command according to the first time derivative of the transient immersion index. The air thermodynamics coordinated control module is used to control the exhaust gas bypass proportional valve to reduce the valve opening according to the wave encounter phase angle when receiving the high sea state coordinated mode signal, and output a lock-up start-on command to the auxiliary scavenging blower. The long-term energy efficiency optimization module is used to calculate the gradient of the energy efficiency cost function composed of specific fuel consumption with respect to the main speed setpoint based on the shaft dynamic parameters and the engine thermodynamic air circuit parameters, and to continuously fine-tune the main speed setpoint according to the descent path of the gradient.

2. The energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion according to claim 1, characterized in that, The multi-source data acquisition module is specifically used for: Extract the raw timestamp of the data packet; Use the engine control system master clock pulse as a unified time alignment reference; Extract the current reference timestamp, and retrieve the adjacent sampled data of the hull kinematic parameters, the shaft dynamic parameters, and the engine thermodynamic gas path parameters on both sides of the current reference timestamp in the data buffer area; Linear interpolation is performed based on the current reference timestamp to output an aligned sequence of synchronization state data.

3. The energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion according to claim 1, characterized in that, The kinematic mapping calculation module is specifically used for: Extract the instantaneous pitch angle and instantaneous heave displacement from the kinematic parameters of the hull; The transient vertical displacement of the propeller shaft relative to the static equilibrium position of the hull is calculated based on the set static parameters of the ship's geometric dimensions, the instantaneous pitch angle, and the instantaneous heave displacement. The instantaneous absolute immersion depth of the propeller is obtained by adding the transient vertical displacement to the still water stern draft. When the instantaneous absolute immersion depth is between zero and twice the effective physical radius of the propeller, the transient proportion of the effective work projection area is derived as the transient immersion index based on the integral calculation model.

4. The energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion according to claim 1, characterized in that, The phase extraction and mode determination module is specifically used for: The transient immersion index is filtered by a high-pass filter to remove the DC bias component and extract the AC disturbance component. The AC disturbance component is input into the built-in phase-locked loop algorithm module to calculate the phase deviation between the AC disturbance component and the reference signal to achieve frequency and phase locking, and output the wave encounter phase angle in real time. The fluctuation amplitude is extracted by using an orthogonal second-order generalized integrator to generate an original signal in phase with the AC disturbance component and an orthogonal signal with a 90-degree lag. The Euclidean norm of the original signal and the orthogonal signal is calculated.

5. The energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion according to claim 1, characterized in that, The phase extraction and mode determination module is also used for: Create a sliding time window of a preset length on the timeline; Within the sliding time window, the root mean square (RMS) value of the fluctuation amplitude is calculated by performing root mean square (RMS) statistics on the fluctuation amplitude. When the root mean square value is determined to be greater than the preset sea state threshold parameter, the high sea state cooperative mode signal is output.

6. The energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion according to claim 1, characterized in that, The fuel mechanical dynamics feedforward module is specifically used for: The fluctuation amplitude is converted into the speed deviation dead zone increment based on the dead zone adaptive gain constant, so as to adjust the control dead zone parameter; The first time derivative of the transient immersion index is compared with a preset negative threshold value; When the first time derivative of the transient immersion index is less than the preset negative threshold, a negative fuel feedforward compensation is generated. Obtain the basic fuel closed-loop command output by the governor, and superimpose the negative fuel feedforward compensation amount with the basic fuel closed-loop command to generate the total fuel injection command.

7. The energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion according to claim 1, characterized in that, The air thermodynamic coordinated control module is specifically used for: Determine whether the wave encounter phase angle falls within the set advance phase angle window; When it is determined that the window has been entered, a turn-off feedforward bias is generated. Obtain the basic gas path valve command output by the gas path controller, and algebraically subtract the basic gas path valve command from the shut-off feedforward offset to control the waste gas bypass proportional valve to reduce the valve opening. Throughout the entire high sea state cooperative mode, the start-up command in a locked state is continuously output to the auxiliary scavenging blower.

8. The energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion according to claim 7, characterized in that, The air thermodynamic coordinated control module is also used for: Real-time acquisition of feedback data on the absolute pressure of the scavenging manifold and the turbocharger speed from the engine thermodynamic air path parameters; The absolute pressure of the scavenging manifold is compared with the set maximum allowable scavenging pressure, and the turbocharger speed feedback data is compared with the set maximum allowable speed. When it is determined that the absolute pressure of the scavenging manifold is greater than the maximum permissible scavenging pressure or the turbocharger speed feedback data is greater than the maximum permissible speed, the algebraic subtraction calculation path between the basic air circuit valve command and the shut-off feedforward bias is forcibly blocked, and the terminal execution command of the exhaust bypass proportional valve is forcibly locked to the maximum safe pressure relief opening.

9. The energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion according to claim 1, characterized in that, The long-term energy efficiency optimization module is specifically used for: The shaft dynamic parameters and engine thermodynamic gas path parameters are smoothed and filtered on a long-period time scale to extract the average fuel consumption mass flow rate and average shaft power under steady-state conditions. The energy efficiency cost function is obtained by dividing the average fuel consumption mass flow rate by the average shaft power. Under the set macro-optimization period, the difference between the energy efficiency cost function and the historical cost function of the previous calculation period is calculated, and the difference is divided by the change in the main speed setting value to calculate the gradient of the energy efficiency cost function with respect to the main speed setting value.

10. The energy efficiency optimization control system for ship propulsion systems based on multi-sensor fusion according to claim 9, characterized in that, The long-term energy efficiency optimization module is also used for: The master rotation speed setting value is fine-tuned according to the set optimization learning rate constant and the gradient. When the absolute value of the gradient in multiple consecutive calculation cycles is less than the preset convergence tolerance threshold, the step fine adjustment of the main speed setting value is paused, the current main speed setting value is locked, and the current average shaft power is recorded as the locking reference value. The system monitors the average shaft power. When it is determined that the average shaft power deviates from the locked reference value by more than the set environmental reset tolerance rate, the system automatically releases the locked state of the main speed setting value and reactivates the step fine-tuning logic.

Images