Robustness design method for active vibration reduction system of marine diesel engine

Abstract

The embodiment of the invention provides a robustness design method for an active vibration reduction system of a marine diesel engine. The robustness design method comprises the steps that 1, a vibration sensor, a rotating speed sensor and an actuator of the active vibration reduction system are installed; 2, identifying a control system; 3, setting an active vibration reduction objective function; 4, acquiring an interference signal d (n); 5, generating a reference signal x (n); 6, designing a filter W (n) of the active vibration reduction system; 7, designing a multi-duration forgetting factor; 8, updating a filter W (n) of the active vibration reduction system in real time, controlling an output signal to drive an execution mechanism to generate excitation, and completing control of a closed loop; 9, monitoring whether the control system fails or not, if yes, executing the step 10, and otherwise, executing the step 11; step 10, starting on-line correction of the line spectrum of the secondary channel; and 11, the control signal is output to an actuator, and active vibration reduction control is completed. According to the embodiment of the invention, the effectiveness of the control effect of the active vibration reduction system is ensured.

Description

Technical Field

[0001] This invention relates to the field of engineering application technology of active vibration and noise reduction technology, and in particular to a robust design method for an active vibration reduction system for marine diesel engines. Background Technology

[0002] Ship vibration and noise control technology has always been a hot topic and focus of research in universities, research institutes, and other institutions. Traditional vibration reduction and isolation measures, represented by rubber isolators and airbags, are insufficient to meet the needs of low-frequency control. Active vibration reduction technology offers greater flexibility and adaptability in controlling low-frequency vibration and noise, and is currently recognized in the industry as an effective method for solving low-frequency vibration problems. Active vibration reduction systems based on the FX-LMS adaptive control algorithm have been widely used and have achieved good control results in laboratory environments. However, in complex engineering applications, the control effect of active vibration reduction is unsatisfactory or even divergent. One of the main reasons is that active vibration reduction systems lack robust design, which restricts the engineering application of active vibration reduction technology.

[0003] In actual ship navigation conditions, the secondary channel of the active vibration damping system changes with the operating state of the ship's own power system and the overall rolling state of the ship. When the secondary channel undergoes unknown changes, the active vibration damping system cannot correct its secondary channel identification parameters in time, which will cause the active vibration damping system to lose its control effect for a period of time. In addition, when the control effect is lost or the vibration level of the ship's propulsion device increases, due to the cancellation mechanism of the vibration control algorithm, the output of the active vibration damping system will continue to increase, causing the control output to reach saturation, resulting in nonlinear output excitation of the control system, affecting the active vibration damping effect, and even causing the control system to diverge. Summary of the Invention

[0004] In view of the above-mentioned problems in the prior art, the present invention provides a robust design method for an active vibration reduction system for marine diesel engines, in order to solve the technical problem that the active vibration reduction control effect fails or even diverges under complex conditions such as changes in the operating state of the power system during actual ship navigation, hull turning and large-amplitude rolling, insufficient output force of the actuator, impact on the control system and sudden changes in error signal quality.

[0005] This invention provides a robust design method for an active vibration reduction system for marine diesel engines, comprising the following steps:

[0006] Step 1: Complete the installation of the vibration sensor, speed sensor and actuator of the active vibration damping system;

[0007] Step 2: Control system identification. Identify the frequency response function of the control channel through the active vibration reduction system.

[0008] Step 3: Set the active vibration reduction objective function. Based on the vibration characteristics of the diesel engine, define the optimal control effect function as Topt;

[0009] Step 4: Acquire interference signal d(n) and acquire vibration sensor signal;

[0010] Step 5: Generate reference signal x(n), acquire the speed sensor signal and convert it into reference signal x(n);

[0011] Step 6: Design the filter W(n) for the active vibration reduction system;

[0012] Step 7: Design a multi-duration forgetting factor based on the size of the filter weight vector N;

[0013] Step 8: The filter W(n) of the active vibration reduction system is updated in real time. The control output signal u(n) = x(n) * W(n) drives the actuator to generate excitation and complete the control closed loop. At this time, the update formula of the error e(n) is e(n) = u(n) + d(n).

[0014] Step 9: Monitor whether the control system has failed. If it has failed, proceed to step 10; otherwise, proceed to step 11.

[0015] Step 10: Start online correction of secondary channel line spectra;

[0016] Step 11: Output the control signal to the actuator to complete the active vibration reduction control.

[0017] In one embodiment, the control system employs an anti-saturation design, specifically including...

[0018] Define the actual control effect function Tp;

[0019] The relationship between the actual control effect function and the optimal control effect function is determined as follows:

[0020] Tp = -Topt / (1+γ)

[0021] Where γ is the output anti-saturation constraint factor;

[0022] Determine the update formula for the corrected error e(k): e(n+1)=(1-1 / γ)u(n)+d(n).

[0023] In one embodiment, the online correction of the secondary channel line spectrum includes,

[0024] The frequency at which the identification control effect fails is defined as ω;

[0025] Based on the assumption that the external interference is a harmonic signal, the expression for the interference signal d*(t) at the failure frequency is determined. in, and These are the fitting coefficients;

[0026] Determine the expression for the failure frequency control signal u(t), u(t) = u c cos(ωt)+u s sin(ωt), where u c and u s These are the fitting coefficients;

[0027] Determine the frequency response function of the control channel The expressions for the real and imaginary parts,

[0028]

[0029] Where ε(t)=e * (t)-e(t), where g is the update iteration step size;

[0030] By obtaining and Replace the original control channel frequency response function The corresponding elements are used to obtain the corrected control channel frequency response function.

[0031] Determine the filter W(n+1) G The updated formula,

[0032]

[0033] Where u is the step size;

[0034] Determine the expression for the control output signal

[0035] U(n) G =X(n)*W(n) G .

[0036] In one embodiment, the weight vector N of the filter satisfies N / C≥25%, where C is the sampling rate of the active vibration reduction system.

[0037] In one embodiment, the identification process in step 2 is as follows:

[0038] Each actuator emits a white noise signal individually, and the transmission relationship between the actuator and each sensor signal is identified by the signals collected by the vibration sensor.

[0039] In one embodiment, the vibration sensor and actuator are both mounted on the engine mounts of the diesel engine and gearbox, and the speed sensor is mounted on the output shaft of the diesel engine.

[0040] In one embodiment, in step 4, the interference signal d(n) is acquired by a data acquisition board.

[0041] In one embodiment, in step 5, the speed sensor signal is acquired by a data acquisition board.

[0042] In one embodiment, the coefficients of the active vibration damping system filter W(n) are an array of size N, and the initial values ​​of the array are generated by random definition.

[0043] In one embodiment, the size of the array N is more than 10 times the control frequency.

[0044] Compared with existing technologies, the beneficial effects of the robust design method for an active vibration damping system for marine diesel engines provided by this invention are as follows: This invention fully recognizes the importance of robust design for active vibration damping systems in real ship environments, and innovatively proposes a multi-pronged active vibration damping system design method that integrates multi-duration forgetting factor design, online correction of control failure caused by offline identification of secondary channels, and a novel low-computation anti-saturation control strategy. This solves the problem of active vibration damping control failure or even divergence under complex conditions such as changes in the operating state of the power system during real ship navigation, hull turning and large-amplitude rolling, insufficient output force of actuators, impact on the control system, and sudden changes in error signal quality. It ensures the effectiveness of the active vibration damping system control effect, realizes the first large-scale application of active vibration damping devices in China, and achieves stable and effective control of the active vibration damping system. Attached Figure Description

[0045] Figure 1 An adaptive control block diagram of Fx-LMS is provided for a robust design method of an active vibration reduction system for a marine diesel engine, as provided in an embodiment of the present invention.

[0046] Figure 2 A schematic diagram of the LMS offline identification algorithm involved in a robust design method for an active vibration reduction system for a marine diesel engine provided in an embodiment of the present invention;

[0047] Figure 3 This is a schematic diagram of the target line spectrum online identification algorithm involved in a robust design method for an active vibration reduction system for marine diesel engines provided in an embodiment of the present invention.

[0048] Figure 4 A schematic diagram of the online correction process for the secondary channel involved in a robust design method for an active vibration reduction system for a marine diesel engine, provided as an embodiment of the present invention;

[0049] Figure 5 This is a schematic diagram of the control system of an active vibration damping system, which is provided as an embodiment of the present invention for a robust design method of an active vibration damping system for a marine diesel engine. Detailed Implementation

[0050] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0051] Various embodiments and features of this application are described herein with reference to the accompanying drawings.

[0052] These and other features of this application will become apparent from the following description of preferred forms of embodiments given as non-limiting examples, with reference to the accompanying drawings.

[0053] It should also be understood that although this application has been described with reference to some specific examples, those skilled in the art can certainly implement many other equivalent forms of this application, which have the features described in the claims and are therefore all within the scope of protection defined herein.

[0054] The above and other aspects, features and advantages of this application will become more apparent when taken in conjunction with the accompanying drawings and in view of the following detailed description.

[0055] Specific embodiments of this application are described below with reference to the accompanying drawings; however, it should be understood that the claimed embodiments are merely examples of this application, which can be implemented in various ways. Well-known and / or repeated functions and structures are not described in detail to ascertain the true intent based on the user's historical operations, and to avoid unnecessary or redundant details that would obscure this application. Therefore, the specific structural and functional details claimed herein are not intended to be limiting, but merely serve as the basis and representative basis for the claims to teach those skilled in the art to use this application in various ways with substantially any suitable detailed structure.

[0056] This specification may use the phrases “in one embodiment,” “in another embodiment,” “in yet another embodiment,” or “in other embodiments,” all of which may refer to one or more of the same or different embodiments according to this application.

[0057] The principles and features of the present invention are described below with reference to the accompanying drawings. The embodiments described are for illustrative purposes only and are not intended to limit the scope of the invention. The following description, in conjunction with... Figure 1-5 The preferred embodiments of the present invention will be described in further detail below:

[0058] like Figure 1-5 As shown in the figure, an embodiment of the present invention provides a robust design method for an active vibration reduction system for marine diesel engines, comprising the following steps:

[0059] Step 1: Complete the installation of the vibration sensor, speed sensor and actuator of the active vibration damping system;

[0060] Step 2: Control system identification. Identify the frequency response function of the control channel through the active vibration reduction system.

[0061] Step 3: Set the active vibration reduction objective function. Based on the vibration characteristics of the diesel engine, define the optimal control effect function as Topt;

[0062] Step 4: Acquire interference signal d(n) and acquire vibration sensor signal;

[0063] Step 5: Generate reference signal x(n), acquire the speed sensor signal and convert it into reference signal x(n);

[0064] Step 6: Design the filter W(n) for the active vibration reduction system;

[0065] Step 7: Design a multi-duration forgetting factor based on the size of the filter weight vector N;

[0066] Step 8: The filter W(n) of the active vibration reduction system is updated in real time. The control output signal u(n) = x(n) * W(n) drives the actuator to generate excitation and complete the control closed loop. At this time, the update formula of the error e(n) is e(n) = u(n) + d(n).

[0067] Step 9: Monitor whether the control system has failed. If it has failed, proceed to step 10; otherwise, proceed to step 11.

[0068] Step 10: Start online correction of secondary channel line spectra;

[0069] Step 11: Output the control signal to the actuator to complete the active vibration reduction control.

[0070] In one embodiment, the control system employs an anti-saturation design, specifically including...

[0071] Define the actual control effect function Tp;

[0072] The relationship between the actual control effect function and the optimal control effect function is determined as follows:

[0073] Tp = -Topt / (1+γ)

[0074] Where γ is the output anti-saturation constraint factor;

[0075] Determine the update formula for the corrected error e(k): e(n+1)=(1-1 / γ)u(n)+d(n).

[0076] In one embodiment, the online correction of the secondary channel line spectrum includes,

[0077] The frequency at which the identification control effect fails is defined as ω;

[0078] Based on the assumption that the external interference is a harmonic signal, the expression for the interference signal d*(t) at the failure frequency is determined. in, and These are the fitting coefficients;

[0079] Determine the expression for the failure frequency control signal u(t), u(t) = u c cos(ωt)+u s sin(ωt), where u c and u s These are the fitting coefficients;

[0080] Determine the frequency response function of the control channel The expressions for the real and imaginary parts,

[0081]

[0082] Where ε(t)=e * (t)-e(t), where g is the update iteration step size;

[0083] By obtaining and Replace the original control channel frequency response function The corresponding elements are used to obtain the corrected control channel frequency response function.

[0084] Determine the filter W(n+1) G The updated formula,

[0085]

[0086] Where u is the step size;

[0087] Determine the expression for the control output signal

[0088] U(n) G =X(n)*W(n) G .

[0089] In one embodiment, the weight vector N of the filter satisfies N / C≥25%, where C is the sampling rate of the active vibration reduction system.

[0090] In one embodiment, the identification process in step 2 is as follows:

[0091] Each actuator emits a white noise signal individually, and the transmission relationship between the actuator and each sensor signal is identified by the signals collected by the vibration sensor.

[0092] In one embodiment, the vibration sensor and actuator are both mounted on the engine mounts of the diesel engine and gearbox, and the speed sensor is mounted on the output shaft of the diesel engine.

[0093] In one embodiment, in step 4, the interference signal d(n) is acquired by a data acquisition board.

[0094] In one embodiment, in step 5, the speed sensor signal is acquired by a data acquisition board.

[0095] In one embodiment, the coefficients of the active vibration damping system filter W(n) are an array of size N, and the initial values ​​of the array are generated by random definition.

[0096] In one embodiment, the size of the array N is more than 10 times the control frequency.

[0097] Example 1

[0098] Based on an in-depth analysis of the technical problems described in the background art, the applicant has found that the failure or even deterioration of the control effect of the above-mentioned active vibration reduction system is due to the abnormal changes in the parameters of the active vibration reduction system caused by changes in the external environment. The applicant recognizes that strengthening the robust design of the active vibration reduction system can solve this engineering application problem.

[0099] The core features of this invention include four aspects: 1) design of multi-duration forgetting factor based on FX-LMS adaptive control algorithm; 2) design of online correction of secondary channel line spectrum; 3) comprehensive robustness design of anti-saturation design; and 4) comprehensive collaborative operation of the above three methods.

[0100] like Figure 5 As shown, the basic principle of the active vibration reduction system is "vibration control with vibration", that is, real-time acquisition of the speed signal and vibration signal of the power equipment related to the controlled characteristic frequency, and processing by the control system according to a certain control law to generate a control signal that is equal in value and opposite in phase to the original vibration signal, thereby canceling the original vibration signal and achieving the purpose of active vibration control.

[0101] The active vibration damping device control system is a multi-input and multi-output control system. The control system reference signal is reconstructed from the speed signal, and the error signal is acquired by the vibration sensor. The controller calculates the above signals in real time, outputs the control signal to the power amplifier, and then performs vibration control through the actuator installed on the propulsion device.

[0102] The physical meanings of the control parameters are explained below:

[0103] W(z) is the adaptive filter parameter, which is the weight array in the control algorithm;

[0104] P(z) is the frequency response function of the interference channel, which is the vibration transmission path of the excitation source of the controlled object (such as a diesel engine), and is an unknown path in the active vibration reduction system;

[0105] H(z) is the frequency response function of the control channel, which is the vibration transmission path from the actuator to the sensor in the active vibration reduction system. In the active control system, it is a known path.

[0106] It is the frequency response function of the control channel being identified, the vibration transmission identification path from the actuator to the sensor in the active vibration reduction system, and H(z) is the mathematical expression of the frequency response function of the control channel.

[0107] x(n) is the reference signal, which can be a directly acquired vibration signal or a signal calculated based on the rotational speed. The conversion algorithm adopts an existing mature algorithm.

[0108] d(n) is the disturbance signal, which is the vibration response signal excited by the controlled object (such as a diesel engine), and can be measured by a sensor;

[0109] u(n) is the control signal, which is the final signal calculated by the entire active vibration reduction algorithm. This signal is actuated by the actuator to generate the vibration response desired by the user.

[0110] The robust design method for an active vibration reduction system for marine diesel engines provided in this invention, based on the FX-LMS adaptive control algorithm, includes the following steps:

[0111] Step 1: Complete the installation of the vibration sensor. (e.g.) Figure 5 As shown, the vibration sensor is mounted on the engine mounts of the diesel engine and gearbox;

[0112] Step 2: Complete the speed sensor installation. As shown in Figure 5, the speed sensor is installed on the output shaft of the diesel engine;

[0113] Step 3: Complete the actuator installation. As shown in step 5, the actuator is installed on the engine mounts of the diesel engine and gearbox;

[0114] Step 4: Control System Identification. This is achieved through identification of the active vibration damping device. The function identification process is as follows: each actuator at each position emits a white noise signal individually, and the transmission relationship between the actuator and each sensor signal is identified by the signal collected by the vibration sensor.

[0115] Step 5: Set the active vibration reduction objective function, that is, to what level the vibration of the diesel engine needs to be controlled based on the vibration characteristics of the diesel engine, and define the optimal function as Topt;

[0116] Step 6: Acquisition of interference signal d(n). The vibration sensor signal is acquired through the acquisition board, which is the interference signal d(n).

[0117] Step 7: Reference signal generation. The speed sensor signal acquired by the acquisition board is converted by the algorithm to become the reference signal x(n).

[0118] Step 8: Design the filter W(n) for the active vibration reduction system. The filter coefficients are initially an array, which can be randomly defined. The size of the array is generally more than 10 times the control frequency, and the array size is defined as N.

[0119] Step 9: Design a multi-duration forgetting factor based on the magnitude of the filter's weight vector N;

[0120] Step 10: The design principle of the multi-duration forgetting factor is that the state of the control system at time N is affected by the combined influence of the previous time 1 to N-1.

[0121] Step 11: The design of the multi-duration forgetting factor includes three aspects: the reference signal X(n), the error signal e(n), and W(n) are the filter weights, and their formulas are shown in Formula 1, Formula 2, and Formula 3.

[0122] Step 12: The filter W(n) of the active vibration damping system is updated in real time, and the control output signal U(n) = X(n) * W(n) drives the actuator to generate excitation, thus completing the control closed loop;

[0123] Step 13: At this time, the vibration of the vibration sensor is the sum of the original vibration d(n) of the diesel engine and gearbox and the vibration U(n) of the actuator, expressed as Formula 4;

[0124] Step 14: To prevent control system overflow, the control system adopts an anti-saturation design;

[0125] Step 15: Define the actual control effect function as Tp;

[0126] Step 16: The relationship between the actual control effect function and the ideal optimal control effect function is as follows:

[0127] Tp = -Topt / (1+γ);

[0128] Step 17: Through anti-saturation design, the expression for the sum of the original vibration d(n) of the diesel engine and gearbox and the actuator vibration U(n) is now corrected to Formula 11;

[0129] Step 18: When there is a deviation in the channel transfer function, such as changes in the structure of the diesel engine or gearbox mounting or changes in boundary constraints, the control system will experience control failure.

[0130] Step 19: At this point, start the online correction algorithm for the secondary channel line spectrum;

[0131] Step 20: Identify the frequency at which the control effect fails, and define it as ω;

[0132] Step 21: The expression for the failure frequency interference signal d*(t) is Equation 5;

[0133] Step 22: The expression for the failure frequency control signal u(t) is Equation 6;

[0134] Step 23: The deviation in the failure frequency identification function (control channel frequency response function) is reflected in the real and imaginary parts of the identification function. The update expressions for the real and imaginary parts are given by Formula 7.

[0135] Step 24: After completing the correction of the failure frequency identification function, replace the original identification function. The corresponding failure frequency identification function elements are used to obtain a modified identification function.

[0136] Step 25: Control channel identification function after online correction become Therefore, the filter W(n+1) G The update formula changes from formula 10 to...

[0137]

[0138] Step 26: Due to the update and change of filter W(n+1), the control output signal changes from Equation 4 to...

[0139] U(n) G =X(n)*W(n) G

[0140] Step 27: Finally, the control signal is output to the actuator to complete the active vibration reduction control.

[0141] in,

[0142] 1) Multi-duration forgetting factor design

[0143] like Figure 1 As shown, in the FX-LMS adaptive control algorithm, let W(z) be the adaptive filter, P(z) be the frequency response function of the interference channel, and H(z) be the frequency response function of the control channel. Here, x(n) is the frequency response function of the identified control channel, x(n) is the input signal at time n (i.e., the reference signal), u(n) is the control signal output by the adaptive controller, d(n) is the interference signal, e(n) is the error signal, and Z(n) is the signal obtained after the reference signal x(n) passes through... The filtered signal.

[0144] Define the reference signal X(n):

[0145] X(n)=[x(n) x(n-1) … x(n-N+1)] Formula 1

[0146] Define the error signal e(n):

[0147] e(n)=[e(n) x(n-1) … e(n-N+1)] Formula 2

[0148] Where: n represents the current sampling time, and the size of the time interval Δt is determined by the sampling rate of the active vibration reduction system. Let the sampling rate of the active vibration reduction system be C.

[0149] W(n) is the filter weight vector, which can be expressed as:

[0150] W(n)=[w0(n)w1(n-1) … w N-1 (n)] Formula 3

[0151] The size of the filter's weight vector is set to N.

[0152] The design of the multi-duration forgetting factor mainly determines the influence of the previous N-1 time moments on the current N time moment through the relationship between the system sampling rate and the filter weight vector. According to the application verification of the active vibration reduction system of marine diesel engines, when N / C≥25%, the active vibration reduction system will not fail or diverge even if it is subjected to multiple impacts in a short period of time.

[0153] 2) Design of online correction for secondary channel line spectra

[0154] The offline identification method simulates the unknown system using an FIR filter of a certain order; its principle block diagram can be seen in [the diagram]. Figure 2 As shown in the figure. P(z) is the unknown disturbance system, and W(z) is the adaptive filter. When the adaptive process converges, W(z) is the best estimate of P(z), and y(n) can best reproduce d(n). At this time, the error signal e(n) is minimized. The adaptive filter and the unknown system are equivalent under this input. The reference input x(n) is broadband white noise.

[0155] Figure 2 Minimize the objective function J = e 2 (n), the gradient descent method is used to identify the digital filter W(z) offline.

[0156] e(n)=x(n)*W(n)+d(n) Formula 4

[0157] The identified filter W(z) is the secondary channel identification parameter, denoted as h(n).

[0158] The online identification method for target line spectra is as follows: Figure 3As shown in the figure, the superscript "*" on some parameters indicates real-time estimated values. The estimated error signal e*(t) is fed back to the controller C, and the controller generates a control signal u(t) that excites the system P, producing a response y(t) to counteract external disturbances d(t). The goal of the control algorithm is to make the system response y(t) infinitely close to -d(t). Assume that the system P is a stationary time-invariant system. Typically, the target line spectrum of the controlled signal d(t) is known. The online identification algorithm needs to calculate the real and imaginary parts of the system P in real time.

[0159] Depend on Figure 3 It can be seen that e(t)=P[u(t)]+d(t), where P[u(t)] represents the time-domain response generated after the control signal u(t) excites the system P.

[0160] Assuming the external interference is a harmonic signal, then d * (t) can be expressed as:

[0161]

[0162] Where ω is the original frequency of the interference signal.

[0163] Under steady-state excitation, the control signal u(t) should be a harmonic signal with the same frequency as the disturbance signal:

[0164] u(t)=u c cos(ωt)+u s sin(ωt) formula 6

[0165] Combining the expression for e(t), under steady-state response, we have:

[0166]

[0167] Among them, P R * With P I * Let J and ε represent the real and imaginary parts of the system's frequency response at frequency ω, respectively. The objective function to be minimized is J = ε. 2 (t) / 2=[e * (t)-e(t)] 2 / 2, Using the gradient descent algorithm, the two unknown parameters are estimated. The recursive form of the estimation formula is as follows:

[0168]

[0169] Here, g represents an update iteration step size, which affects the search speed of the algorithm. It is usually taken as a small value to ensure the stability of the algorithm.

[0170] When a control failure occurs during the active control process of the target line spectrum, the parameter P is identified online using the target line spectrum. R * With P I * The offline identification parameter h(n) is corrected, and the corrected h * (n) This is then fed into the active control system for effective control of the controlled object. The flowchart for online correction of secondary channel offline identification is as follows: Figure 4 As shown.

[0171] 3) Anti-saturation design

[0172] In an active vibration reduction system, the performance index of the control system is only related to the square of the error signal. When the output power of the control system cannot meet the control requirements, the optimal array solution obtained according to the performance index, denoted as Wopt (hereinafter the same), is always below the current solution of the system, denoted as Wp (hereinafter the same). When the control filter's Wp causes the control system output to reach the maximum output threshold uα, but the optimal performance index state Wopt has not yet been reached, the residual error of the control system will still drive the current solution Wp of the control filter to continuously update towards the optimal solution Wopt. In the subsequent update process, the output u(k) of the control system is already greater than the output threshold u(α) of the controller, and the control signal is clipped, causing nonlinearity in the system. The anti-saturation control strategy proposed in this patent modifies the system performance related to the error signal. According to the actual control requirements of the active vibration reduction system, the system performance modification weight can be flexibly adjusted, ultimately making the new performance target array solution of the control system approach Wp, thereby ensuring that the control output is within the maximum threshold uα.

[0173] The derived formula is as follows:

[0174] Let the new optimal performance index be:

[0175] minJ2=E{e 2 (k)+γu 2 Formula 8 (k)

[0176] In the formula, γ is called the output anti-saturation constraint factor.

[0177] At this point, the optimal solution T of the control algorithm opt and the current solution T p The relationship becomes:

[0178] Tp=-Topt / (1+γ) Formula 9

[0179] Therefore, it can be seen that after reconstructing the performance indicators of the control system, this patent makes the current calculated solution of the filter proportional to the optimal solution of the system. By adjusting the output anti-saturation constraint factor, the proportional relationship between the current solution and the optimal solution can be adjusted, thereby suppressing the saturation output of the control system.

[0180] The iterative formula for the weighting coefficients is obtained by replacing J2 with the instantaneous performance index:

[0181]

[0182] The update formula for the error e(k) is as follows:

[0183]

[0184] Compared to the classic anti-saturation control strategy based on the Sigmoid function, the Sigmoid function adds a high-order exponential operation, which greatly increases the computational load of the control algorithm. The novel anti-saturation control strategy proposed in this embodiment of the invention achieves the purpose of anti-saturation of the control output by reconstructing the performance of the control system through simple addition / subtraction operations, and has the typical characteristics of low computational load.

[0185] The embodiments of this invention are robust designs for active vibration reduction systems using multiple parallel design algorithms. By employing a multi-pronged approach, the robustness of the active vibration reduction system is ensured from multiple angles. This robust design method has been effectively applied in marine diesel engines, guaranteeing the effectiveness of the active vibration reduction device's control effect under various conditions, including changes in the operating state of the ship's own power system, hull turning and large-amplitude rolling, insufficient output force of the actuator, impacts on the control system, and sudden changes in error signal quality.

[0186] The above embodiments are merely exemplary embodiments of the present invention and are not intended to limit the present invention. The scope of protection of the present invention is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to the present invention within its spirit and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of the present invention.

Claims

1. A robust design method for an active vibration reduction system for marine diesel engines, characterized in that, Includes the following steps: Step 1: Complete the installation of the vibration sensor, speed sensor and actuator of the active vibration damping system; Step 2: Control system identification. Identify the frequency response function of the control channel through the active vibration reduction system. Step 3: Set the active vibration reduction objective function. Based on the vibration characteristics of the diesel engine, define the optimal control effect function as Topt; Step 4: Acquire interference signal d(n) and acquire vibration sensor signal; Step 5: Generate reference signal x(n), acquire the speed sensor signal and convert it into reference signal x(n); Step 6: Design the filter W(n) for the active vibration reduction system; Step 7: Design a multi-duration forgetting factor based on the size of the filter weight vector N; Step 8: The filter W(n) of the active vibration reduction system is updated in real time. The control output signal u(n) = x(n) * W(n) drives the actuator to generate excitation and complete the control closed loop. At this time, the update formula of the error e(n) is e(n) = u(n) + d(n). Step 9: Monitor whether the control system has failed. If it has failed, proceed to step 10; otherwise, proceed to step 11. Step 10: Start online correction of secondary channel line spectra; Step 11: Output the control signal to the actuator to complete the active vibration reduction control.

2. The robust design method for an active vibration reduction system for marine diesel engines according to claim 1, characterized in that: The control system employs an anti-saturation design, specifically including: Define the actual control effect function Tp; The relationship between the actual control effect function and the optimal control effect function is determined as follows: Tp = -Topt / (1+γ) Where γ is the output anti-saturation constraint factor; Determine the update formula for the corrected error e(k): e(n+1)=(1-1 / γ)u(n)+d(n).

3. The robust design method for an active vibration reduction system for marine diesel engines according to claim 2, characterized in that: The online correction of the secondary channel line spectrum includes, The frequency at which the identification control effect fails is defined as ω; Based on the assumption that the external interference is a harmonic signal, the expression for the interference signal d*(t) at the failure frequency is determined. in, and These are the fitting coefficients; Determine the expression for the failure frequency control signal u(t), u(t) = u c cos(ωt)+u s sin(ωt), where u c and u s These are the fitting coefficients; Determine the frequency response function of the control channel The expressions for the real and imaginary parts, Where ε(t)=e * (t)-e(t), where g is the update iteration step size; By obtaining and Replace the original control channel frequency response function The corresponding elements are used to obtain the corrected control channel frequency response function. Determine the filter W(n+1) G The updated formula, Where u is the step size; Determine the expression for the control output signal U(n) G =X(n)*W(n) G 。 4. The robust design method for an active vibration reduction system for marine diesel engines according to claim 1, characterized in that: The weight vector N of the filter satisfies N / C≥25%, where C is the sampling rate of the active vibration reduction system.

5. The robust design method for an active vibration reduction system for marine diesel engines according to claim 1, characterized in that: The identification process in step 2 is as follows: Each actuator emits a white noise signal individually, and the transmission relationship between the actuator and each sensor signal is identified by the signals collected by the vibration sensor.

6. The robust design method for an active vibration reduction system for marine diesel engines according to claim 1, characterized in that: The vibration sensor and actuator are both mounted on the engine mounts of the diesel engine and gearbox, and the speed sensor is mounted on the output shaft of the diesel engine.

7. The robust design method for an active vibration reduction system for marine diesel engines according to claim 1, characterized in that: In step 4, the interference signal d(n) is acquired by the acquisition board.

8. The robust design method for an active vibration reduction system for a marine diesel engine according to claim 1, characterized in that: In step 5, the speed sensor signal is acquired by the acquisition board.

9. The robust design method for an active vibration reduction system for a marine diesel engine according to claim 1, characterized in that: The coefficients of the active vibration reduction system filter W(n) are an array of size N, and the initial values ​​of the array are generated by random definition.

10. The robust design method for an active vibration reduction system for a marine diesel engine according to claim 1, characterized in that: The size of the array N is more than 10 times the control frequency.

Images