Pressure control method and system for marine ammonia fuel supply system based on model-free adaptive control

Abstract

The invention discloses a pressure control method and system for a marine ammonia fuel supply system based on model-free self-adaptive control, and the method employs a model-free self-adaptive control scheme based on tight format dynamic linearization, and carries out the real-time collection of system pressure and valve control quantity data. And dynamically estimating a pseudo partial derivative phi (k) on line to capture system characteristics, and designing an adaptive control law to calculate a valve opening instruction according to the system characteristics. A dual-robust mechanism is formed by introducing an estimation weight factor mu and a control weight factor lambda, and noise interference and controlled quantity mutation are effectively suppressed. The method can automatically adapt to system dynamic change and external disturbance, high-precision and strong-robustness stable control over the ammonia fuel supply pressure is achieved under the complex ship working condition, and the safety and reliability of the ammonia power ship are remarkably improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of ship power control system, and particularly relates to a pressure control method and system for a marine ammonia fuel supply system based on model-free adaptive control (MFAC), which is suitable for a ship engine fuel supply system using ammonia as single fuel or main fuel (such as ammonia-diesel dual fuel). BACKGROUND

[0002] With the urgent demand for green and low-carbon transformation of global shipping industry, zero-carbon or low-carbon fuels represented by ammonia are gradually becoming one of the key paths for decarbonization of ship power systems due to their mature storage and transportation, suitable energy density, and no carbon dioxide emission in the combustion process. The commercial application of ammonia fuel engines (including pure ammonia engines and ammonia-diesel dual fuel engines) puts forward higher control requirements for the fuel supply system. Among them, the accurate and stable control of ammonia fuel supply pressure is the core link to ensure the efficient, safe and clean operation of the engine, and directly affects the combustion efficiency, emission performance, system reliability and dynamic response quality.

[0003] At present, the pressure control of marine fuel supply systems (including traditional fuel systems and new fuel systems such as liquefied natural gas (LNG) and methanol) generally adopts methods based on classical control theory, with proportional-integral-derivative control (PID) and its derivative structures (such as cascade PID and feedforward-PID compound control) being the mainstream. Such methods have a certain control effect in application scenarios where the working conditions are relatively stable and the dynamic characteristics of the system change little. However, marine ammonia fuel supply systems have characteristics such as nonlinearity, time-varying and difficulty in modeling, and their dynamic behavior is affected by factors such as liquid ammonia properties, pipeline resistance, valve characteristics, load fluctuations, and ship navigation conditions (such as roll and pitch).

[0004] The existing PID-based control methods or advanced control strategies relying on accurate mechanism models generally have poor adaptability, insufficient robustness, tedious parameter tuning, and difficulty in maintaining the accuracy and stability of pressure control in a wide range of working conditions when facing such complex objects. The above limitations not only restrict the operating performance and safety and reliability of ammonia fuel power ships, but also hinder the development process of energy cleaning and low-carbonization of the shipping industry.

[0005] Therefore, it is urgent to develop a new type of control method that does not rely on accurate mathematical models of the system, can adapt to the dynamic changes and external disturbances of the system, has strong robustness, and is easy to implement in engineering, to effectively solve the key technical problems in the pressure control of marine ammonia fuel supply. SUMMARY

[0006] The purpose of this invention is to overcome the inherent defects of existing marine ammonia fuel supply pressure control methods, such as strong dependence on mathematical models, poor adaptability in nonlinear time-varying systems, weak robustness, and complex parameter tuning. This invention provides a pressure control method and system for marine ammonia fuel supply systems based on MFAC, which utilizes only the real-time input and output data of the system to achieve online self-tuning and self-adaptation of controller parameters. This enables high-precision and robust control of ammonia fuel supply pressure under complex and variable operating conditions of ships, while reducing the complexity of engineering applications.

[0007] To achieve the above-mentioned objectives, the technical solution adopted by the present invention is as follows: On the one hand, the present invention provides a pressure control method for a marine ammonia fuel supply system based on model-free adaptive control, characterized by comprising the following steps: S1. Set the target pressure value: Set the target pressure P at the outlet of the ammonia fuel supply system based on the real-time load demand of the engine. ref ( k ), and send it to the control module, where k For discrete-time indexing; S2. Acquire system I / O data: Real-time acquisition of current data. k The actual output pressure value P of the ammonia fuel supply system k ), and the previous moment k -1 Control input quantity applied to the actuator u ( k -1); S3. Calculate the pressure tracking error e( at the current moment). k The formula is as follows: e( k ) = P ref ( k ) - P( k ) S4. Calculate the change, including the pressure change ΔP between adjacent time points. ), the change in control input Δu at adjacent time points k -1); S5. Constructing an online pseudo-partial derivative estimation algorithm: Using the current and historical I / O data {P(k), P(k-1), ..., u(k-1), u(k-2), ...} collected in step S2, the pseudo-partial derivative φ(k), which characterizes the time-varying gain of the system, is dynamically captured and updated in real time through an online pseudo-partial derivative estimation algorithm; the estimation algorithm is as follows: φ(k) =φ(k-1)+[η×ΔP(k)-φ(k-1)×Δu(k-1)] ×Δu(k-1) / [μ+|Δu(k-1)|^2] Where: ΔP(k)=P(k)-P(k-1) is the pressure change between adjacent time steps; Δu(k-1)=u(k-1)-u(k-2) is the control input change between adjacent time steps; η∈(0,1] is the step size factor, which controls the convergence speed of the estimation; μ>0 is the weighting factor; Furthermore, constraints are imposed on the estimated value φ(k) to ensure that φ min ≤φ(k)≤φ max , where φ min >0; S6. Design a model-free adaptive control law: based on the estimated φ(k) and pressure tracking error e(k). k ), calculate the control input at the current time k using the following formula: u(k) = u(k-1) + [ρ×φ(k)×e(k)] / [λ+|φ(k)|^2] Where: ρ ∈(0,1] is the step size factor of the controller, which affects the strength of the control action and the convergence speed; λ>0 is the weight factor of the controller, which is used to limit the change range of the control quantity and balance the tracking performance and the smoothness of the control input; The calculated control input u(k) is subjected to amplitude limiting to ensure it conforms to the actual physical constraints of the actuator [u min , u max ].

[0008] S7. Output control command and execute: Output the calculated and limited control input u(k) to the actuator to adjust the high-pressure pump frequency in real time, thereby changing the ammonia fuel flow rate and realizing closed-loop control of the supply pressure.

[0009] S8. Online rolling update: At the next sampling time k+1, the above steps are repeated to achieve continuous adaptive update of control parameters and continuous optimization of control action.

[0010] Furthermore, the estimated weighting factor μ>0 is used to prevent numerical instability caused by an excessively small denominator in the estimation algorithm when the system input changes gradually, and to suppress the interference of measurement noise on the pseudo-partial derivative estimation.

[0011] Furthermore, the control weight factor λ>0 is used to balance the speed of pressure tracking with the smoothness of control action, and to prevent drastic changes in control commands due to sudden changes in pseudo-partial derivative φ(k) or excessive tracking error e(k).

[0012] In another aspect, the present application also provides a ship ammonia fuel power control system, comprising a controller configured to perform ammonia fuel supply pressure control, the controller comprising a processor and a memory, characterized in that the memory stores a computer program which, when executed by the processor, configures the controller to be specifically used for: Real-time acquisition and processing of pressure sensor data, calculation of pressure tracking error; Based on the tight format dynamic linearization method, only relying on the input and output data stream of the system, the dynamic estimation and update of the pseudo partial derivative φ(k) are performed online; Taking the real-time updated φ(k) as the core adaptive parameter, the control command is dynamically generated through the model-free adaptive control law; The control command is limited and output to the actuator; So as to realize the above-mentioned method.

[0013] Compared with the prior art, the present application has the following beneficial effects: 1) Abandoning the dependence on the complex nonlinear mechanism model of the ammonia fuel supply system of the ship, only using the I / O data (pressure value and control quantity) of the real-time operation of the system for control. By online estimating the key parameter of pseudo partial derivative (PPD), the real-time characteristics of the system are dynamically captured and tracked, which fundamentally solves the core pain point of the control performance in this field due to the difficulty in accurate modeling.

[0014] 2) The tight format dynamic linearization (CFDL) technology is adopted, and the essentially nonlinear ammonia fuel supply system is equivalent to a time-varying linear data model at each dynamic working point. Double robustness design is introduced: weight factor is introduced in the PPD estimation algorithm, which effectively suppresses measurement noise interference and ensures numerical stability; under complex operating conditions such as ship starting, acceleration, deceleration, steady state and variable sea conditions, the target pressure can be quickly and accurately tracked, and the steady-state error and dynamic deviation are significantly reduced. It has strong inhibition ability to internal parameter perturbation of the system (such as valve characteristic change, pipeline resistance change), unmodeled dynamics and external strong disturbance (such as engine load step change, ship shaking). It can automatically adapt to the slow change (such as equipment aging) and fast change (such as working condition switching) of the dynamic characteristics of the system, without the need for frequent manual parameter re-tuning. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is the flowchart of the pressure control method of the ammonia fuel supply system of the ship based on MFAC of the present application. DETAILED DESCRIPTION

[0016] The present application will be further described in detail below in combination with the drawings and examples, but should not limit the protection scope of the present application.

[0017] Please refer toFigure 1 , Figure 1 is a flow chart of the pressure control method of the MFAC-based marine ammonia fuel supply system of the present application, as shown, a model-free adaptive control method for marine ammonia fuel supply pressure is applied to a marine ammonia fuel supply system containing a liquid ammonia storage tank, a low-pressure pump, a high-pressure pump, a buffer tank, a frequency converter, and a pressure sensor. The method is based on the model-free adaptive control (MFAC) theory of compact form dynamic linearization (CFDL), the core is to linearize the nonlinear system near the dynamic operating point, and only use the online I / O data of the system to update the controller parameters in real time. The specific steps include: Set the pressure target value: according to the real-time load demand of the engine and the safety operation requirement, determine the target pressure value P_ref(k) of the ammonia fuel supply system outlet, wherein k is the discrete time index.

[0018] Collect system I / O data: real-time acquisition of ammonia fuel supply system actual output pressure value P(k) at current time k (measured by pressure sensor) and control input quantity u(k-1) (such as valve opening degree instruction or valve position signal) applied to actuator (pressure regulating valve) at last time k-1.

[0019] Calculate the pressure tracking error: e(k) = P_ref(k) - P(k).

[0020] Construct a pseudo partial derivative (PPD) online estimation algorithm: use the current and historical I / O data of the system {P(k), P(k-1),..., u(k-1), u(k-2),...} to estimate the key parameter of the system dynamic characteristic-pseudo partial derivative (Pseudo Partial Derivative, φ(k)) in real time.

[0021] The PPD parameter φ(k) reflects the sensitivity (i.e. equivalent gain) of the system output (pressure P) to the change of input (control quantity u) near the current operating point, and the core form of its estimation formula is: φ(k) = φ(k-1) + [η × ΔP(k) - φ(k-1) × Δu(k-1)] × Δu(k-1) / [μ + |Δu(k-1)|^2] Wherein: ΔP(k) = P(k) - P(k-1) is the pressure change between adjacent time points.

[0022] Δu(k-1) = u(k-1) - u(k-2) is the control input change between adjacent time points.

[0023] η ∈ (0,1] is a step factor, which controls the convergence speed of the estimation.

[0024] μ>0 is a weight factor, which is used to enhance the robustness of the estimation algorithm and prevent numerical instability caused by too small denominator.

[0025] The estimated value φ(k) is subjected to the constraint |φ(k)|>ε or projected to the preset interval [φ_min, φ_max], to ensure its boundedness and reasonableness (ε is a small positive number to prevent division by zero error).

[0026] The model-free adaptive control law is designed: based on the estimated PPD φ(k) and the pressure tracking error e(k), the control input u(k) at the current time k is calculated: u(k) = u(k-1) + [ρ × φ(k) × e(k)] / [λ + |φ(k)|^2] Wherein: ρ ∈ (0,1] is a step factor of the controller, which affects the strength and convergence speed of the control action.

[0027] λ>0 is a weight factor of the controller, which is used to limit the variation amplitude of the control input, to balance the tracking performance and the smoothness of the control input (suppress overshoot and oscillation).

[0028] The calculated control input u(k) is subjected to amplitude limiting processing, to make it comply with the actual physical constraints [u_min, u_max] of the actuator (frequency 0%-100%).

[0029] The control instruction is output and executed: the calculated and amplitude-limited control input u(k) is output to the actuator, to adjust the high-pressure pump frequency in real time, thereby changing the ammonia fuel flow, to realize closed-loop control of the supply pressure.

[0030] Online rolling update: at the next sampling time k+1, the above steps are repeated to realize continuous adaptive update of the control parameters and continuous optimization of the control action.

[0031] The specific embodiments of the present application are described above. It should be understood that the present application is not limited to the above specific embodiments, and those skilled in the art can make various modifications or modifications within the scope of the claims, which does not affect the essential content of the present application. The above embodiments and features in the embodiments can be combined with each other without conflict.

Claims

1. A pressure control method for a marine ammonia fuel supply system based on model-free adaptive control, characterized by, Comprising the following steps: S1. Set pressure target value: set target pressure P of ammonia fuel supply system outlet according to engine real-time load demand ref ( k ), and send to control module, wherein k is a discrete time index; S2. Collect system I / O data: real-time collect the current time k 's ammonia fuel supply system actual output pressure value P( k ), and the last time k -1 applied to the actuator control input u ( k -1) S3. Calculate the current time pressure tracking error e( k ), the formula is as follows: e( k ) = P ref ( k ) - P( k ) S4. Calculate the variation, including the pressure variation ΔP( ) of the adjacent time, the control input variation Δu( k -1) of the adjacent time; S5. Constructing pseudo partial derivative online estimation algorithm: using the current and historical I / O data {P(k), P(k-1),..., u(k-1), u(k-2),...} collected in step S2, the pseudo partial derivative φ(k) representing the time-varying gain of the system is dynamically captured and updated in real time through the pseudo partial derivative online estimation algorithm; the estimation algorithm is: φ(k) = φ(k-1) + [η × ΔP(k) - φ(k-1) × Δu(k-1)] × Δu(k-1) / [μ + |Δu(k-1)|^2] Where: ΔP(k) = P(k) - P(k-1) is the pressure change between adjacent time points; Δu(k-1) = u(k-1) - u(k-2) is the control input change between adjacent time points; η ∈ (0, 1] is the step factor, which controls the convergence speed of the estimation; μ > 0 is the weight factor; And, a constraint is imposed on the estimated value φ(k) so that it satisfies φ min ≤φ(k)≤φ max where φ min > 0; S6. Design model-free adaptive control law: Based on the estimated φ(k) and the pressure tracking error e(k), the control input quantity at the current time k is calculated, as follows: k u(k) = u(k - 1) + Kp e(k) + Kφ φ(k) u(k) = u(k-1) + [ρ × φ(k) × e(k)] / [λ + |φ(k)|^2] Where: ρ ∈ (0, 1] is the step factor of the controller, which affects the strength and convergence speed of the control action; λ > 0 is the weight factor of the controller, which is used to limit the change amplitude of the control quantity and balance the tracking performance and the smoothness of the control input; The calculated control input u(k) is subjected to a clipping process to make it comply with the actual physical constraints [u min , u max ] of the actuator.

2. S7. Output control command and execute: output the calculated and limited control input quantity u(k) to the actuator to adjust the frequency of the high-pressure pump in real time, thereby changing the ammonia fuel flow and realizing closed-loop control of the supply pressure.

3. S8. Online rolling update: repeat the above steps at the next sampling time k+1 to realize continuous adaptive update of the control parameters and continuous optimization of the control action.

4. The pressure control method of the marine ammonia fuel supply system according to claim 1, characterized by, The estimation weight factor μ > 0 is used to prevent numerical instability caused by too small denominator in the estimation algorithm when the system input changes slowly, and to suppress the interference of measurement noise on the pseudo partial derivative estimation.

5. The method of pressure control of a marine ammonia fuel supply system according to claim 1, characterized in that, The control weight factor λ > 0 is used to balance the rapidity of pressure tracking and the smoothness of control action, and to prevent the control command from changing dramatically due to sudden change of the pseudo partial derivative φ(k) or too large tracking error e(k).

6. A marine ammonia fuel power control system comprising a controller configured to perform ammonia fuel supply pressure control, the controller comprising a processor and a memory, characterised in that, The memory stores a computer program, which, when executed by the processor, configures the controller to specifically: Real-time acquisition and processing of pressure sensor data to calculate the pressure tracking error; Based on the compact format dynamic linearization method, only relying on the input and output data stream of the system, the dynamic estimation and update of the pseudo partial derivative φ(k) are performed online; Using the real-time updated φ(k) as the core adaptive parameter, the control command is dynamically generated through the model-free adaptive control law; The control command is limited and output to the actuator; Thus, the method of any one of claims 1 to 5 is realized.

Images