A dynamic inertia adaptive control method for diesel storage of ship low-frequency pulse load

Abstract

This invention provides a diesel-storage dynamic inertia adaptive control method for low-frequency pulse loads on ships. The method includes obtaining a functional relationship between the power imbalance and the JVSG (Junctional Power System Generation) based on a virtual synchronous generator, diesel generator, pulse load model, and system power imbalance, thereby obtaining the output power of the DG and VSG (Dynamic Power Generation Generator) during the entire pulse switching process. Then, a time-domain explicit function for the power allocation between the VSG and DG is generated. Based on the power imbalance and JVSG... VSG From the functional relationship and the explicit function in the time domain, the VSG virtual inertia J is obtained. VSG The influence of the value characteristics on the unbalanced power at different stages of the system; based on J VSG The impact of the value characteristics on the unbalanced power of the system at different stages is investigated. A pulse-based adaptive strategy is employed, introducing adaptive coefficients and tuning these coefficients. This invention offers a faster response to disturbances, demonstrates superior performance in reducing system unbalanced power, and is more feasible.

Description

Technical Field

[0001] This invention relates to the field of energy storage control, and in particular to a dynamic inertia adaptive control method for diesel-electric storage under low-frequency pulse loads on ships. Background Technology

[0002] Due to significant differences in response capabilities between energy storage systems and generators, during pulsed load switching, the two cannot achieve instantaneous power distribution according to capacity ratios. In this situation, either the diesel generator or the energy storage system may temporarily bear excessive active power, leading to transient overcurrent problems. To improve the power synchronization capability between the diesel generator and the energy storage system, the energy storage control strategy needs to be optimized. However, in actual operation, due to the influence of the diesel generator set, generator shaft system, and control loop, the precise value of the diesel generator's shaft rotational inertia is uncertain, making it difficult to accurately match the virtual inertia of the energy storage converter with the diesel generator's inertia under pulsed load conditions. This poses challenges to the design and optimization of the energy storage control strategy.

[0003] Currently, energy storage converters typically employ droop control to simulate the external characteristics of diesel generators, dynamically adjusting the frequency and amplitude of the output voltage based on measured power to autonomously participate in system power distribution and frequency regulation. When using a Virtual Synchronous Generator (VSG) control strategy, the system can not only simulate the static external characteristics of a synchronous generator but also reproduce its dynamic characteristics such as rotational inertia. Modern diesel-energy storage hybrid power supply systems often use VSG control strategies for their energy storage converters. Current research has proposed various control strategies to address power oscillation and unbalanced power issues in diesel-energy storage parallel operation systems, mainly including inertia matching and dynamic damping enhancement. The key challenges and shortcomings of similar technologies are mainly concentrated in the following aspects: 1) In the face of the power imbalance and transient circulating current problem of diesel-storage power system caused by high-power low-frequency pulse load, existing research has proposed an adaptive energy management strategy to adjust the power allocation of the system according to the pulse load demand, energy storage SOC and voltage. However, this method only achieves the balanced allocation of system power in steady state and does not improve the transient active circulating current of the system. 2) To address the difficulty in determining the inertia parameters of diesel generators, existing research proposes a coordinated control strategy for additional torque. This strategy aims to reduce the angular acceleration difference in the diesel-storage hybrid power supply system by dynamically adjusting the virtual torque of the energy storage converter. However, the calculation of the additional torque involves designing the power differential term, and the controller has a latching time, which has limitations and results in torque adjustment lag. This significantly impacts the actual unbalanced power suppression effect. Summary of the Invention

[0004] The purpose of this invention is to provide a diesel-storage dynamic inertia adaptive control method for low-frequency pulse loads on ships, which has a faster response speed to disturbances, a better actual effect on reducing system unbalanced power, and higher feasibility.

[0005] To achieve the above objectives, this invention provides a diesel-powered fuel cell dynamic inertia adaptive control method for low-frequency pulse loads on ships, comprising the following steps: Step 1: Based on the virtual synchronous generator VSG, diesel generator DG, pulse load model and system power imbalance, obtain the functional relationship between power imbalance and JVSG, and thus obtain the output power of DG and VSG respectively during the entire pulse switching process; Step 2: Based on the output power of DG and VSG during the entire pulse switching process, obtain the time-domain explicit function of power distribution between VSG and DG; based on the power imbalance and J... VSG From the functional relationship and the explicit function in the time domain, the VSG virtual inertia J is obtained. VSG The influence of the value characteristics on the unbalanced power of the system at different stages, including the 0~DT stage and the DT~T stage; Step 3, based on J VSG The influence of the value characteristics on the unbalanced power at different stages of the system is investigated. An adaptive strategy for the entire pulse process is introduced, and the adaptive coefficient is tuned to limit the range of the VSG virtual inertia change rate.

[0006] Preferably, the power imbalance is related to J VSG Functional relationship: (1); ; (2); The parameters in the formula are as follows: (3); In the formula, P dc (s) represents the DC-side power, L(s) represents the transfer function between the system power imbalance and the equivalent pulse power, and k DG k VSG J represents the synchronization power coefficient of DG and VSG, respectively. DG J is the moment of inertia of the diesel generator shaft system. VSG Here, s represents the VSG virtual inertia, and s is a complex frequency domain variable. For VSG virtual damping, S is the damping coefficient of the diesel generator. DG and S VSG These are the rated capacities of DG and VSG, respectively. System rated angular velocity.

[0007] The preferred output power equations for DG and VSG throughout the entire process are as follows: (4); In the formula, L(s) represents the unbalanced power.

[0008] Preferably, the time-domain explicit function for VSG and DG power allocation is: (5); In the formula, P dc L(t) represents the DC-side power, and L(t) represents the unbalanced power. This refers to the output power of the diesel engine. This refers to the output power of the VSG.

[0009] Preferably, during the 0~DT phase, after the system has basically completed power distribution, we have: (6); in, , These respectively represent the characteristics of VSG inertia values. , The magnitude of the unbalanced power of the system after the pulse load power distribution is completed in the 0~DT phase.

[0010] Preferably, in the DT~T stage, after power allocation is completed, we have: (7); in, , These respectively represent the characteristics of VSG inertia values. , At that time, the magnitude of the unbalanced power of the system after the pulse load power distribution is completed in the DT~T phase.

[0011] The preferred final equation for the fully adaptive strategy is: (8); in, m is the maximum virtual inertia of VSG inc m is the inertia increase factor for the 0~DT stage. dec Here, denoted as the inertia decrease factor for the DT~T stage, where T is the period and D is the duty cycle. The initial virtual inertia of the VSG, with adaptive coefficients m inc and m dec Used to control the rate of change of inertia during the pulse loading and cutting phases, respectively: (9).

[0012] Preferred, m inc mdec The range of values: (10); In the formula, K d The frequency regulation effect coefficient of the load is represented. P is the damping ratio of a second-order system. L Where f is the pulse power and f is the pulse frequency. K p This is the active power droop coefficient.

[0013] Therefore, the present invention employs the aforementioned adaptive control method for dynamic inertia of diesel-electric storage for low-frequency pulse loads on ships, and its technical effects are as follows: This invention establishes a mathematical model of the relationship between the power imbalance of a diesel-storage hybrid power supply system for low-frequency pulse loads on ships and the virtual inertia of the VSG. Based on this mathematical model, the time-domain equations of the output power of the DG and VSG are analyzed.

[0014] This invention analyzes the influence of the virtual inertia value characteristics on the magnitude of system unbalanced power during the entire process of pulse load switching, and proposes a VSG dynamic inertia adaptive control method that tracks pulse load parameters.

[0015] The control method proposed in this invention ensures that the VSG inertia adapts to the duty cycle and period of the pulse load, dynamically matching the shaft rotation inertia of the diesel generator, thereby suppressing the unbalanced power fundamental wave and its odd harmonic components.

[0016] This invention takes into account the millisecond-level delay in pulse parameter acquisition and pulse parameter identification errors in actual engineering. The dynamic inertia adaptive control method has strong anti-interference ability and can achieve optimization effect in such scenarios, thus having practical applicability.

[0017] Compared with other existing control strategies, this invention has a faster response speed to disturbances, a more effective reduction of unbalanced power in the system, and is more feasible. Attached Figure Description

[0018] Figure 1 Topology for shipboard diesel-storage hybrid power supply system; Figure 2 This is a diesel generator model; Figure 3 For VSG active frequency control; Figure 4 The waveform in the time domain is a pulse load. Figure 5 The root locus diagram of the system; Figure 6 The amplitude-frequency response of power imbalance under different JVSG values; Figure 7 Analysis of the relationship between unbalanced power and inertia; Figure 7 (a) is the unbalanced power when the VSG virtual inertia is less than the DG inertia; Figure 7 (b) is the unbalanced power when the VSG virtual inertia is greater than the DG inertia; Figure 8 The effect is for a pulse condition with T=200ms and D=50%; Figure 8 (a) represents the change in virtual inertia; Figure 8 (b) represents the change in power imbalance; Figure 8 (c) represents the active circulating current; Figure 8 (d) represents pulse power, diesel engine power, and VSG power; Figure 9 A comparison between normal load and pulse load conditions; Figure 9 (a) Power curves for conventional load and pulse load; Figure 9 (b) Power imbalance caused by conventional load and pulsed load; Figure 10 for T =250ms, D =40% pulse condition effect; Figure 10 (a) represents the change in virtual inertia; Figure 10 (b) represents the change in power imbalance; Figure 10 (c) represents the active circulating current; Figure 10 (d) represents pulse power, diesel engine power, and VSG power; Figure 11 This refers to the triangular pulse wave and its corresponding effect on system power imbalance. Figure 11 (a) represents the power imbalance; Figure 11 (b) is the active circulation; Figure 12 A cluster of amplitude-frequency response curves; Figure 13 Comparison of system transient active power circulation outside and inside the parameter tuning range; Figure 14 The active power circulating current of the system corresponding to different parameter values; Figure 15 This method is compared with existing research methods; Figure 15 (a) represents the power imbalance; Figure 15 (b) is the active circulation; Figure 16 To take into account the control effect of actual pulse parameter acquisition delay; Figure 16 (a) represents virtual inertia; Figure 16 (b) represents the power imbalance. Figure 17 To verify the effectiveness of the strategy considering the identification bias of actual parameters. Detailed Implementation

[0019] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0020] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0021] Example 1 A method for adaptive dynamic inertia control of diesel-electric storage for low-frequency pulse loads on ships includes the following steps: Step 1: Diesel-Storage Hybrid Power Supply System for Low-Frequency Pulse Loads Based on the establishment of models for a virtual synchronous generator (VSG), a diesel generator (DG), and a pulsed load, the relationship between the power imbalance of the transient process system and the VSG inertia J is derived. VSG Pulse load pulse power P L By establishing the functional relationship between duty cycle D and period T, the output power equations for DG and VSG during the entire pulse switching process are obtained.

[0022] Step 1.1, Diesel-Storage Hybrid Power Supply System like Figure 1 The diagram shows the topology of a shipboard diesel-electric hybrid power supply system with low-frequency pulse loads. The diesel generator mainly includes a prime mover, a synchronous generator, and a control system. The energy storage system includes battery energy storage, an energy storage converter, and a corresponding control system. Both systems work together to compensate for power fluctuations caused by sudden load changes. The energy storage converter uses VSG control. The main components of the load side are high-power pulse loads and conventional loads. Figure 1 In the middle, L DG and L VSG Indicates the inductance from the DG and VSG output ports to the AC bus, i DG i VSG C represents the output current of DG and VSG, respectively. dc For DC-side voltage regulation capacitor, U dc This indicates the DC side voltage.

[0023] Diesel generator control model such as Figure 2 As shown. The speed regulation time constant of the generator is usually on the order of seconds, which is much larger than the dynamic response time of VSG based on virtual inertial control (on the order of milliseconds). Therefore, the speed regulation link of the diesel generator in the diesel-storage hybrid power supply system of this invention can be approximated as a static link and ignored.

[0024] The rotor motion equation of a diesel generator reflects the mechanical motion characteristics of the rotor during transient processes: (1); Among them, J DG P is the moment of inertia of the diesel generator shaft system. DGn and P DGe D represents the rated power and electromagnetic power of the diesel generator. DG This refers to the damping coefficient of the diesel generator. and These represent the system's rated angular velocity and the generator's actual angular velocity, respectively.

[0025] The VSG control loop mainly consists of two parts: a droop control loop and a voltage-current double-closed loop. The droop control loop is composed of an active-frequency control loop, and the control equation is as follows: (2); Among them, P VSGm P ref These are the VSG virtual mechanical power and active reference value, respectively, K p This is the active power droop coefficient. This refers to the angular velocity of the VSG virtual rotor.

[0026] VSG rotor motion equation: (3); J VSG The energy for the VSG virtual inertia is provided by the energy storage side, D VSG P is the VSG damping coefficient. VSGe This indicates the electromagnetic power output of the VSG. Its active frequency control is as follows: Figure 3 As shown.

[0027] The pulsed load power mainly consists of two parts: steady-state power and dynamic pulse power, and its waveform is as follows: Figure 4 As shown, dynamic pulse power affects the transient processes of the diesel-storage hybrid power system, and its main characteristic parameter is the pulse power P. L The pulse period T and duty cycle D are defined as follows: the pulse period reflects the frequency of pulse load changes, the pulse power reflects the maximum impact amplitude of the pulse load, and the pulse duty cycle reflects the duration of the peak power of the pulse load.

[0028] P p Represents peak power, typically (4~7)P. s P p P sMaintain a constant inertia. Generally, the period T of a low-frequency pulse load is greater than or equal to 100ms. Conventional loads have small power amplitudes and slow changes, and commonly used droop control and constant inertia control methods can effectively suppress the power imbalance of the system. For the aforementioned low-frequency pulse loads, due to their low-frequency peaks, complex inertia matching, and high dynamic response requirements, the inertia control is significantly more difficult than that of conventional loads.

[0029] In the time domain, the dynamic pulse power equation is: (4); In the formula, n is the number of periods.

[0030] DC side capacitor C dc During pulse load switching, the DC-side voltage can be kept stable through charging and discharging, compensating for the power compensation gap on the AC side. The DC-side power is denoted as P. dc As the equivalent pulse power, we have: (5); In the formula, It is a DC voltage.

[0031] Performing a Laplace transform on it, the equivalent pulse power is obtained in the complex frequency domain: (6); In the formula, s represents a complex frequency domain variable.

[0032] In the diesel-storage hybrid power system, the change in pulse power is shared by the diesel generator and the VSG, therefore the transient process always has the following power balance equation: (7); In the formula, This refers to the output power of the diesel engine.

[0033] Step 1.2, Unbalanced Power of Diesel Storage For DG and VSG, let their rated capacities be S respectively. DG and S VSG Ideally, the goal of system power allocation is to distribute the output power of DG and VSG according to their capacities S. DG / S VSG Distribution. During pulse load switching, due to the precise inertia parameter J of the diesel generator set... DG The external situation presents an uncertain state, VSG virtual inertia J VSG With J DG During dynamic processes, direct matching is impossible, causing a deviation between the actual power distribution ratio and its capacity ratio. The system power imbalance is defined as: (8); Based on the modeling of DG and VSG using equations (1) and (3), the angular accelerations of DG and VSG in the complex frequency domain can be obtained: (9); in, , For the angular accelerations of DG and VSG, For VSG virtual damping, For diesel engine damping, This refers to the mechanical input power of the diesel engine.

[0034] By analogy with the power angle characteristics of DG, the output electromagnetic power of DG and VSG can be approximately expressed as: (10); in, For diesel engine power angle, For VSG power angle, U DG U VSG These are the output port voltages of DG and VSG, respectively, U bus X is the AC bus voltage. DG X VSG k represents the reactance between the output ports of DG and VSG and the AC bus, respectively. DG k VSG These represent the synchronization power coefficients of DG and VSG, respectively: (11); Substituting equations (9) and (10) into equation (8), and performing Laplace transform and coefficient homogenization, we have: (12); From this, we can obtain the power imbalance and J. VSG Functional relationship: (13); The parameters in the formula are as follows: (14); make: (15); L(s) represents the transfer function between the system power imbalance and the equivalent pulse power. The essence of active circulating current is the current distribution deviation after the normalization of the DG and VSG capacities, and its change is positively correlated with the system power imbalance. Therefore, suppressing... It can also reduce the system's circulating current.

[0035] By combining equations (7), (8), and (13), we can obtain the output power equations for DG and VSG throughout the entire process: (16); Stability analysis was performed on the third-order system L(s), and the root locus diagram of the system was plotted as follows: Figure 5 As shown, with J VSG When the value of is dynamically increased (from 0.2 to 50) and all other system parameters remain unchanged, the eigenvalues ​​of the third-order system move closer to the imaginary axis and are all located in the left half of the coordinate plane, thus the system can maintain stability.

[0036] Step 2: Correlation Analysis of Virtual Inertia and Unbalanced Power Based on the mathematical model established in step 1, by analyzing the amplitude-frequency curve of the system power imbalance and the power output equation, J is obtained. VSG The influence of the value characteristics on the unbalanced power of the system is further explained by J. VSG How to dynamically match the shaft rotational inertia J of a diesel generator? DG Lay the theoretical foundation.

[0037] According to equation (13), the amplitude-frequency response curve of the power imbalance is plotted as follows: Figure 6 As shown, the power imbalance and VSG virtual inertia J are obtained. VSG Relationship of value characteristics: J VSG Increasing the value of J can reduce the power imbalance of the system. Using the pulse load frequency as the fundamental frequency, the fundamental wave and its odd harmonics of the unbalanced power are suppressed to some extent, but the suppression effect is not significant. To achieve the ideal suppression effect on the system power imbalance, considering the entire process of pulse loading and unloading, J... VSG The method of dynamically following the changes in pulse load is used for segmented optimization.

[0038] The following section discusses the power imbalance of the diesel-storage hybrid power system and J during the entire process of pulse load loading and unloading. VSG The relationship between the values ​​of L(s) is analyzed. Using the Cardano formula, the characteristic equation of L(s) in equation (15) is solved, and its poles p1, p2, and p3 are obtained as follows: (17); In the formula, q is the unbalanced power oscillation factor, and θ is the oscillation power phase angle offset coefficient.

[0039] in: (18); In the formula, r is the unbalanced power unsteady-state factor.

[0040] Based on the poles, the time-domain response of L(t) can be further expressed as a superposition of three modes: (19); t is time, p i Let be the output power of the i-th machine.

[0041] For VSG and DG, performing an inverse Laplace transform on equation (16) yields the explicit time-domain functions of their power distribution: (20); By adjusting J VSG The value of will change the pole distribution of L(t), further affecting the power allocation between the VSG and DG. The time-domain curves of both are shown below. Figure 7 As shown: The blue and red curves represent the VSG inertia value characteristics, respectively. , The output power curve is shown, with the black curve representing the output power of the DG under the corresponding conditions.

[0042] contrast Figure 7 (a) and Figure 7 (b) The following is a detailed analysis of the impact of the VSG inertia value characteristics on the magnitude of the system's unbalanced power during the entire pulse load switching process: During the 0~DT phase, after the system has basically completed power distribution, we have: (twenty one); in, , These respectively represent the characteristics of VSG inertia values. , At that time, the magnitude of the unbalanced power of the system after the pulse load power distribution is completed in the 0~DT phase. Combined with... Figure 6 Conclusion: J VSG Dynamic variation of J is more effective at suppressing system power imbalance than taking a constant value; therefore, J is more effective at this stage. VSG The dynamic value should be incremented to simultaneously reduce the unbalanced power in both the transient process and the power distribution phase. Therefore, in the 0~DT phase, the VSG inertia dynamic process should adopt a dynamically increasing strategy.

[0043] During the DT~T phase, after power allocation is completed, we have: (twenty two); in, , These respectively represent the characteristics of VSG inertia values. , The unbalanced power of the system after the pulsed load power distribution is completed in the DT~T phase is considered. Analogous to the analysis in the pulsed load loading phase, the VSG inertia value characteristic should be decreasing. Therefore, in the DT~T phase, the VSG inertia dynamic process should adopt a dynamically decreasing strategy.

[0044] Combination formula (20) and Figure 7 It can be seen that the design of the above adaptive strategy logic is essentially to adjust the power imbalance of the system by changing the L(t) term. Extending this to atypical rectangular pulse loads (with a certain ramp-up delay during switching), P... dc A small change in (t) will not affect J VSG The value characteristics of the value affect the power imbalance of the system in the 0~DT and DT~T stages, and further, the universality of the adaptive strategy logic design for pulse loads can be obtained.

[0045] Step 3: Adaptive Control Method for Dynamic Inertia of Diesel Storage This step presents the equations expressing the inertia adaptive strategy of the VSG. Its design logic is that the inertia adaptively increases from 0 to DT and adaptively decreases from DT to T. Based on the active frequency model of the VSG and the pulse load parameters, the range of values ​​for the adaptive coefficients is limited.

[0046] Step 3.1, Dynamic Inertia Design During the pulse loading phase, increasing the virtual inertia can reduce the power response speed of the VSG to the pulse load, making its power change closer to that of the DG. This can be achieved through J... VSG With slope m inc The unbalanced power of the system is gradually increased and reduced during the 0~DT phase. Therefore, during pulse loading, the virtual inertia is designed with the following adaptive strategy: (twenty three); Where, m inc The inertia increase factor for the 0~DT stage. This represents the VSG virtual inertia value at the pulse loading moment. J VSG The peak value reached during the inertia increase phase is denoted as J, as the pulse cut-out time VSG The value of . The mathematical model established in this invention. Based on the complex frequency domain, the final result needs to be verified by the model. Therefore, the formula in the loading stage is transformed into the complex frequency domain as follows: (twenty four); During the pulse load cut-off phase, J VSG Adaptive reduction can decrease the system power imbalance after power allocation between VSG and DG: (25); (26); In the formula, m is the maximum virtual inertia of VSG decLet be the inertia reduction coefficient for the DT~T stage. Rearranging the above pulse loading and unloading inertia adaptive strategy in the complex frequency domain, we obtain the final equation for the whole-process adaptive strategy: (27); in, The initial virtual inertia of the VSG, with adaptive coefficients m inc and m dec Used to control the rate of change of inertia during the pulse loading and cutting phases, respectively: (28); Step 3.2, Adaptive parameter tuning The following is about m inc and m dec Parameter tuning is performed to limit the range of the VSG virtual inertia rate of change. Its small-signal model can be described using a second-order system as follows: (29); Where d is the differential symbol, K represents the change in VSG power angle. d The frequency regulation effect coefficient represents the load.

[0047] make: (30); The damping ratio of this second-order system can be obtained as follows: (31); For a second-order VSG system, its damping ratio should be set within a reasonable range to avoid underdamped oscillations or overdamped, sluggish response. Let the range of its damping ratio be... Then there is (32); Get J VSG The range of values: (33); In addition, P p With P s The larger the difference, the greater the pulse power P. L The larger the peak value of J, the better. VSG The rate of change needs to be increased accordingly, and the two are positively correlated. In the control implementation, m inc and m dec The upper limit of the value can be determined by the pulse peak value P. L Further limitations are imposed. Combining the amplitude-frequency response curve's suppression effect on the fundamental frequency and its harmonic components of unbalanced power, m is obtained. inc m dec The range of values: (34).

[0048] K p K is the active droop factor, f is the pulse frequency, and K is the active droop factor. d The frequency regulation effect coefficient of the load is represented. Let be the damping ratio of the second-order system.

[0049] Simulation verification To evaluate the effectiveness of the proposed adaptive inertia control strategy, based on Figure 1 The ship's diesel-storage hybrid electric power system topology shown is simulated in Matlab / Simulink. The simulation parameters are shown in Table 1 below. For ease of analysis, J in subsequent simulations... VSG All values ​​have been normalized to a dimensionless form: (35); The value of VSG inertia before normalization is expressed in kg·m. 2 .

[0050] Table 1 Simulation Parameters

[0051] Scenario 1: Adaptive Policy Verification For T=200ms, D=50%, P L A pulse load of 40kW is applied at 2s. J VSG Adaptive strategy, power imbalance, transient active circulating current, and power output of each power source in the system after adopting the strategy, such as... Figure 8 As shown in Table 2, the JVSG value for the constant inertia method is 0.2.

[0052] Simulation results show that the VSG adopts the proposed adaptive strategy based on pulse parameters, which significantly reduces the power imbalance and transient active circulating current between heterogeneous power sources during pulse load switching. The circulating current amplitude is reduced by 3.4A during pulse loading and by 4A during pulse unloading, demonstrating significant effectiveness.

[0053] Table 2 Comparison of Constant Inertia and Adaptive Methods

[0054] Scenario 2: Adaptability verification under different pulse loads The start-up, shutdown, or operating condition switching time of conventional loads is typically on the order of seconds. For low-frequency pulse loads, the peak power is reached instantaneously during loading / unloading or operating condition switching, followed by continuous high-power surges with a period on the order of milliseconds. Compared to the inertia control of conventional loads, the main challenge lies in the instantaneous high power and the continuous periodic power surges.

[0055] First, a simulation comparison of the system power imbalance of the constant inertia control method under conventional load and pulsed load conditions is conducted, such as... Figure 9 As shown, the continuous periodic power surges of a pulsed load cause continuous power oscillations in the system. In contrast, for conventional loads, the system responds quickly to power imbalances, has a short recovery time, and exhibits smaller oscillation amplitudes under the same conditions. Pulsed loads, however, are characterized by low-frequency peaks, complex inertia matching, and high dynamic response requirements.

[0056] The effectiveness of the proposed strategy for pulse loads with different parameters was verified by changing the period and duty cycle of the pulse load. A pulse load of T=250ms, D=40%, and P were selected. L =40kW, loaded at 2s, the simulation results are as follows. Figure 10 As shown, the power imbalance is significantly reduced, the system circulating current amplitude is reduced by 3.6A during the pulse loading phase and by 4.8A during the cut-out phase, and the power output of the DG and VSG has good synchronization.

[0057] The effectiveness of this dynamic inertia adaptive control method under different pulse load types is verified below, taking a triangular wave pulse as an example. T=200ms, D=50%, P... L =46kW, loaded at 2s, pulse load waveform, and the effect of system power imbalance before and after the method are as follows: Figure 11 As shown.

[0058] To achieve quantitative analysis of simulation results and measure the optimization effect of system power imbalance within a single pulse cycle, the following definition is made: (36); This represents the reduction in system power imbalance before and after adopting the strategy. It serves as a numerical indicator to measure the optimization effect. The larger the value, the more significant the effect of the control method on reducing unbalanced power.

[0059] For triangular wave pulses, simulation results show that the reduction in system power imbalance is significantly improved during the pulse loading phase, while the suppression effect decreases during the cutoff phase. Considering the entire pulse load switching process, the performance index... For reference, the optimized values ​​are basically consistent with those under typical rectangular pulse waves, which further illustrates that the control method proposed in this paper is adaptable to different types of pulse loads.

[0060] To verify the effectiveness of the proposed strategy, this method was tested multiple times with different pulsed load parameters. The pulse power was controlled at 40kW, the pulse period ranged from 100ms to 300ms, and the pulse duty cycle ranged from 30% to 70%. The optimization results are shown in Table 3 above. The results show that the lower the pulse frequency, the more significant the reduction in unbalanced power. Furthermore, under the same frequency band of pulsed load, the optimization effect of this method is less affected by the pulse duty cycle.

[0061] Table 3. 40kW Pulse Load

[0062] The following verifies the superiority of the adaptive strategy proposed in this invention compared to VSG's use of a constant inertia. For example... Figure 12 As shown, in comparison with J VSG The amplitude-frequency characteristics of the system power imbalance were obtained with values ​​of 0.2, 0.5, 1, 1.5, and 1.8, and their dynamic changes. The results show that the inertia adaptive strategy proposed in this invention is significantly better than the case where the inertia changes to a constant value in reducing the amplitude of the system power imbalance.

[0063] Scenario 3: Validation of Adaptive Parameter Tuning Effectiveness To verify m inc m dec The validity of parameter tuning is determined by setting the pulse load parameter to P. L =40kW, T=250ms, D=40%, under this operating condition, substituting the specific parameter values ​​into the formula, we obtain the limiting range of the adaptive coefficient: (37); Select two sets of data, one outside and one inside the parameter's tuning range, and compare m. inc =50,m dec =33.34 and m inc =25, m dec =16.67 The circulating current between the two machines under two operating conditions, such as Figure 13 As shown, outside the parameter tuning range, the adaptive strategy will produce a significant delay in tracking the pulse load, with the maximum transient active circulating current being around 9A, which is further increased compared to the constant inertia strategy, thus contradicting the optimization objective.

[0064] Within the parameter tuning range, different parameter values ​​result in varying degrees of optimization for system power distribution. The adaptive coefficients are set to the following four groups: m inc =20、m dec =13.34, m inc =25、m dec =16.67, m inc=30、m dec =20, m inc =35、m dec =23.3, and the actual optimization results of the system are shown in Table 4 below.

[0065] Table 4. Optimization effect of different parameter values ​​within the parameter tuning range

[0066] Analysis of the data in the table above shows that, within the parameter tuning range, the optimization effect exhibits a trend of first increasing and then decreasing. The adaptive coefficients with the highest optimization degree for system power allocation are concentrated in m. inc =30, m dec =Around 20. Simulation results of cyclic current fluctuations corresponding to different adaptive coefficient values ​​are as follows: Figure 14 As shown, it can be verified that the parameter value is within m. inc =30, m dec =The optimization effect is most obvious around 20.

[0067] For atypical rectangular pulse loads, the optimization effect of different parameter values ​​within the parameter tuning range remains consistent with that in Table 4 above. This is because the trend is mathematically reflected in L(t) with respect to m. inc m dec The partial derivative of the pulse load type does not change due to the fine-tuning of the pulse load type.

[0068] Scenario 4: Comparison with other methods and simulation of real-world scenarios Existing research has achieved extended lifespan and efficient energy utilization of energy storage systems through SOC balancing, but it has not significantly improved the system's power imbalance and transient active power circulation. Furthermore, some scholars have proposed adding torque to VSG control to reduce system imbalance power. However, in practical applications, the calculation of the added torque and the controller latching time lead to slow control response, causing the actual compensation effect for imbalance power to deviate from the expected result. In contrast, the inertia adaptive strategy of this method only needs to be implemented based on given pulse parameters, achieving a fast response. The following is a comparison of the actual simulation results of this patent and the methods used in existing research. Figure 15 As shown.

[0069] Furthermore, in practical engineering applications, there is a 1-10ms delay in obtaining pulse load parameters. To verify the robustness of the proposed dynamic inertia control method under actual action delays, the following example uses a 10ms response delay, with pulse load T=200ms, D=50%, and P... L =40kW, loaded at 2s, and started operating at 2.01s.

[0070] like Figure 16As shown, comparing the response delay control with the ideal zero-delay control reveals that: in the 0~DT stage, the control method's effect on suppressing system power imbalance is weakened under the delay condition, with a peak reduction of 0.015; while in the DT~T stage, the response delay control's effect on suppressing power imbalance is better than the ideal zero-delay state. Therefore, considering the entire process analysis, the parameter acquisition delay has a very small impact on the optimization effect of this dynamic inertia adaptive control method, and this method has strong practical applicability.

[0071] In some harsh operating environments of actual shipboard integrated power systems, pulse load parameters may change in real time, making it difficult to accurately obtain P. L The parameters D and T are then used. At this point, the system's parameter identification device is needed: by real-time monitoring of the AC bus power or frequency changes, the key characteristics of the recently occurred pulse can be quickly identified. When a pulse edge (rising or falling edge) is detected, its period T and duty cycle D are estimated by measuring its power amplitude rise / fall time and combining this with observations of previous pulses. However, this parameter identification has a certain degree of bias. The robustness of this strategy when parameter deviations (±10%) are verified below.

[0072] With actual pulse load T=220ms, D=50%, P L Taking a 40kW pulsed load as an example, consider a scenario with significant deviation in practice: the identified pulse parameters are T=220ms, D=45%, P... L =40kW, with a delay of 1ms, plot the power imbalance curves of the system under conditions of no parameter identification error, considering parameter identification error, and fixed virtual inertia. Figure 17 As shown.

[0073] As can be seen from the above simulation, even if there is a deviation in the pulse parameter identification, the dynamic matching method can still suppress the power imbalance of the system with very little impact. This verifies that the dynamic inertia adaptive control method has strong robustness and practicality.

[0074] Therefore, the present invention adopts the above-mentioned adaptive control method of diesel-storage dynamic inertia for low-frequency pulse loads on ships, which has a faster response speed to disturbances, a better actual effect on reducing unbalanced power of the system, and higher feasibility.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for adaptive control of diesel-powered fuel cell dynamic inertia for low-frequency pulse loads on ships, characterized in that, Includes the following steps: Step 1: Based on the virtual synchronous generator VSG, diesel generator DG, pulse load model and system power imbalance, obtain the functional relationship between power imbalance and JVSG, and thus obtain the output power of DG and VSG respectively during the entire pulse switching process; Step 2, based on the respective output power of the DG and the VSG in the whole process of pulse switching, the time-domain explicit function of the power distribution of the VSG and the DG is obtained; based on the function relationship between the power imbalance degree and J VSG , the time-domain explicit function, the value characteristics of the virtual inertia J VSG of the VSG are obtained, which affect the imbalance power in different stages of the system, including the 0~DT stage, the DT~T stage. Step 3, based on J VSG The influence of the value characteristics on the unbalanced power in different stages of the system, the adaptive strategy throughout the pulse process, the introduction of the adaptive coefficient, the parameter setting of the adaptive coefficient, and the limitation of the range of the virtual inertia change rate of the VSG.

2. The adaptive control method for dynamic inertia of diesel-powered fuel cells for low-frequency pulse loads on ships according to claim 1, characterized in that, Power imbalance and J VSG Functional relationship: （1）； ； （2）； The parameters in the formula are as follows: （3）； In the formula, P dc (s) represents the DC-side power, L(s) represents the transfer function between the system power imbalance and the equivalent pulse power, and k DG k VSG J represents the synchronization power coefficient of DG and VSG, respectively. DG J is the moment of inertia of the diesel generator shaft system. VSG Here, s represents the VSG virtual inertia, and s is a complex frequency domain variable. For the virtual damping of VSG, S is the damping coefficient of the diesel generator. DG and S VSG These are the rated capacities of DG and VSG, respectively. System rated angular velocity.

3. The adaptive control method for dynamic inertia of diesel-powered fuel cells for low-frequency pulse loads on ships according to claim 1, characterized in that, The output power equations for DG and VSG throughout the entire process are as follows: （4）； In the formula, L(s) represents the unbalanced power.

4. The adaptive control method for dynamic inertia of diesel-powered fuel cells for low-frequency pulse loads on ships according to claim 1, characterized in that, The time-domain explicit function of VSG and DG power allocation: （5）； In the formula, P dc L(t) represents DC power, and L(t) represents unbalanced power. This refers to the output power of the diesel generator. This refers to the output power of the diesel generator.

5. The adaptive control method for dynamic inertia of diesel-powered fuel cells for low-frequency pulse loads on ships according to claim 1, characterized in that, During the 0~DT phase, after the system has basically completed power distribution, we have: （6）； in, , These respectively represent the characteristics of VSG inertia values. , The magnitude of the unbalanced power of the system after the pulse load power distribution is completed in the 0~DT phase.

6. The adaptive control method for dynamic inertia of diesel-powered fuel cells for low-frequency pulse loads on ships according to claim 1, characterized in that, During the DT~T phase, after power allocation is completed, we have: （7）； in, , These respectively represent the characteristics of VSG inertia values. , At that time, the magnitude of the unbalanced power of the system after the pulse load power distribution is completed in the DT~T phase.

7. The adaptive control method for dynamic inertia of diesel-powered fuel cells for low-frequency pulse loads on ships according to claim 1, characterized in that, The final equation of the fully adaptive strategy: （8）； in, m is the maximum virtual inertia of VSG inc m is the inertia increase factor for the 0~DT stage. dec Here, denoted as the inertia decrease factor for the DT~T stage, where T is the period and D is the duty cycle. The initial virtual inertia of the VSG, with adaptive coefficients m inc and m dec Used to control the rate of change of inertia during the pulse loading and cutting phases, respectively: （9）。 8. The adaptive control method for dynamic inertia of diesel-powered fuel cells for low-frequency pulse loads on ships according to claim 1, characterized in that, m inc m dec The range of values: （10）； In the formula, K d The frequency regulation effect coefficient of the load is represented. P is the damping ratio of a second-order system. L Where f is the pulse power and f is the pulse frequency. K p This is the active power droop coefficient.

Images