Intelligent self-adaptive control method for diesel generating set

Abstract

The invention discloses an intelligent self-adaptive control method for a diesel generating set, which aims at the problems of slow load fluctuation response, poor system stability and the like of the traditional diesel generating set, and comprises the following steps: triggering a quick response mechanism by monitoring a load power change rate in real time, and dividing a load into low, medium and high load areas by adopting a piecewise linearization power distribution strategy; matching an optimal power combination in a power distribution table generated offline by combining an energy storage SOC state; the winding current is decoupled and controlled through rotor flux linkage directional coordinate transformation, the electromagnetic torque is rapidly adjusted through a PI controller, and cooperative power output is achieved in cooperation with space vector pulse width modulation. The preset rule limits that the energy storage power does not exceed 80% rated value and the SOC is maintained at 20%-80%, and the power distribution table is stored in the digital signal processor for real-time execution. According to the method, the response time can be shortened, the controller load is reduced, and the system stability and the fuel economy under complex working conditions are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of coordinated control of diesel generator sets and energy storage systems, and particularly relates to an intelligent adaptive control method for diesel generator sets. Background Art

[0002] In the prior art, there are significant technical bottlenecks in the power control method of diesel generator sets. Taking the Chinese patent application CN119561445A as an example, it discloses an adaptive power control method for a dual-winding induction generator system, which decouples the control winding current through rotor flux-oriented coordinate transformation and uses a dynamic programming algorithm to allocate the power between the generator and the energy storage unit. However, this method has the following defects: 1. Insufficient dynamic response ability: The dynamic programming algorithm of the prior art is based on offline optimization and needs to pre-calculate the power distribution strategy in the future time period. When the load undergoes sudden fluctuations (such as short-term impact loads or fast-starting and stopping equipment), the system cannot adjust the power distribution in real time, resulting in an instantaneous drop in the output voltage / frequency of the generator, which may trigger protection actions and affect the load stability. 2. Contradiction between algorithm complexity and real-time performance: The dynamic programming algorithm needs to traverse multiple layers of state space (time step, energy storage state, power distribution), and the calculation amount is huge. The prior art lacks real-time optimization measures and is difficult to meet the millisecond-level control response requirements in industrial scenarios. If relying on a high-performance computing platform, it will significantly increase the hardware cost and deployment difficulty.

[0003] The above defects limit the flexibility and reliability of diesel generator sets under complex working conditions, and there is an urgent need for an intelligent control scheme that takes into account both the dynamic response speed and real-time performance. Summary of the Invention

[0004] An object of the present invention is to solve at least the above defects and provide at least the advantages described hereinafter.

[0005] The present invention provides an intelligent adaptive control method for a diesel generator set, which shortens the power regulation response time, reduces the controller calculation load, and improves the operation stability and fuel economy of the system under complex working conditions through real-time power change rate monitoring and piecewise linearized power distribution strategy.

[0006] An intelligent adaptive control method for a diesel generator set provided by the present invention includes

[0007] including the following steps:

[0008] Step 1) Decouple the control winding current of the dual-winding induction generator into an exciting current component and a torque current component through rotor flux-oriented coordinate transformation;

[0009] Step 2) Real-time collect the load power fluctuation signal and calculate the power change rate, and trigger a fast response mechanism when the power change rate exceeds the set threshold;

[0010] Step 3) Based on the piecewise linear power distribution strategy according to preset rules, match the optimal generator output power value and the charge and discharge power value of the energy storage unit in the power distribution table generated offline according to the load power change rate and the current state of charge (SOC) of the energy storage unit;

[0011] Step 4) Adjust the control winding current component through a proportional-integral controller to achieve rapid adjustment of the generator electromagnetic torque;

[0012] Step 5) Use space vector pulse width modulation technology to generate a driving signal for the static excitation converter to complete the coordinated power output of the generator and the energy storage unit;

[0013] The preset rules include:

[0014] The piecewise linearization rule that divides the load power range into a low load area, a medium load area, and a high load area,

[0015] The constraint rule that limits the charge and discharge power of the energy storage unit not to exceed 80% of the maximum power and maintains the state of charge between 20% - 80%;

[0016] The power distribution table is pre-generated through offline simulation and stored in the controller, and the controller uses a digital signal processor to achieve real-time call and execution of the power distribution strategy.

[0017] Preferably, step 2) includes:

[0018] Step 2.1: Synchronously collect the three-phase voltage and current signals of the load using a current transformer and a voltage sensor;

[0019] Step 2.2: Calculate the real-time values of active power and reactive power based on the instantaneous power theory;

[0020] Step 2.3: Perform a moving average process on the power values of consecutive N sampling periods, and calculate the power change rate ΔP / Δt = [P(k) - P(k - N)] / (N·Ts), where Ts is the sampling period (unit: second, usually taking Ts = 1 / 10000 second, i.e., 100 μs, to meet the millisecond-level response requirement); where, P(k) is the instantaneous power value at the kth sampling period (unit: kW or MW), P(k - N) is the instantaneous power value at the k - Nth sampling period (the same unit as P(k)), and N is the number of consecutive sampling periods (a positive integer, such as N = 10 representing the most recent 10 periods);

[0021] Step 2.4: When the absolute value of ΔP / Δt exceeds the preset power change rate threshold range of ±5% rated power / millisecond, immediately trigger the fast response mechanism.

[0022] Step 2.5: The fast response mechanism includes: switching to the high-frequency PID control mode, increasing the regulation rate of the control winding current to 300% of the rated value, and starting the power compensation program of the energy storage unit.

[0023] Preferably, step 3) includes:

[0024] Step 3.1: Determine the load interval (low load area, medium load area or high load area) where the current load power is located according to the load power change rate and the preset piecewise linearization rule;

[0025] Step 3.2: Determine the allowable range of the charge and discharge power of the energy storage unit according to the current state of charge (SOC) of the energy storage unit and the preset state of charge constraint rule;

[0026] Step 3.3: In the power distribution table generated offline, match the generator output power distribution coefficient and the energy storage charge and discharge power distribution coefficient corresponding to the current load interval and the allowable range of the energy storage charge and discharge power;

[0027] Step 3.4: Calculate the target output power value of the generator and the target charge and discharge power value of the energy storage unit at the current moment according to the matched distribution coefficients;

[0028] Step 3.5: Verify whether the calculated generator output power and the energy storage charge and discharge power meet the preset power constraint conditions. If not, return to step 3.1 to rematch;

[0029] Step 3.6: When all constraint conditions are met, output the matched generator output power value and the energy storage charge and discharge power value as the optimal solution.

[0030] Preferably, step 4) includes:

[0031] Step 4.1: Establish a double closed-loop control structure with the torque control loop as the outer loop and the current control loop as the inner loop (objective: when dynamically responding to load changes, quickly adjust the power output through the torque loop). For the multi-closed-loop nested design, it includes the outer power loop → the middle torque loop → the inner current loop. Step 1.4 is activated during system startup or steady state to ensure the stability of the magnetic flux; step 4.1 is triggered during load mutation or power regulation demand and takes over the control right; the seamless connection between the two is achieved through the mode switching logic (such as the power change rate threshold); that is, step 1.4 is the basic field-oriented control, which needs to stabilize the rotor magnetic flux (the outer-loop excitation current, equivalent to the outer power loop) first to provide a stable basis for subsequent torque control. Step 4.1 is the dynamic power regulation, which needs to directly respond to load changes (the outer-loop torque, equivalent to the middle torque loop) and quickly adjust the output through the current loop. Step 1.4 belongs to the bottom-layer current decoupling control, and step 4.1 belongs to the upper-layer power coordination control. The two belong to different control stages;

[0032] Step 4.2: Calculate the reference value of the control winding current component according to the target output power value of the generator and the target charge-discharge power value of the energy storage unit output in Step 3 through the power-current mapping relationship;

[0033] Step 4.3: Adjust the control winding current component by using a digital PI controller. The PI controller parameters are set as follows: proportional coefficient KP = 2.5, integral coefficient KI = 0.8, and sampling period T = 500 μs;

[0034] Step 4.4: When performing closed-loop adjustment on the control winding current component, set the current loop bandwidth to 200 Hz, the adjustment period to 1 ms, and the output limit value to 150% of the rated current;

[0035] Step 4.5: When it is detected that the output of the integral term exceeds 80% of the limit value, start the integral separation mechanism to temporarily remove the integral effect and prevent integral saturation;

[0036] Step 4.6: Real-time collect the feedback value of the control winding current component through a current sensor, compare it with the reference value calculated in Step 4.2, and generate a current adjustment error signal;

[0037] Step 4.7: Input the current adjustment error signal into the PI controller and output a control winding voltage modulation signal;

[0038] Step 4.8: Convert the voltage modulation signal into drive pulses through SVPWM technology to control the power switch devices of the static excitation converter.

[0039] Preferably, Step 5) includes:

[0040] Step 5.1: Perform Clarke transformation on the voltage modulation signal output in Step 4 to convert the voltage components in the three-phase static coordinate system into the two-phase static αβ coordinate system;

[0041] Step 5.2: Perform Park transformation according to the rotor flux orientation angle to convert the voltage components in the αβ coordinate system to the dq rotating coordinate system;

[0042] Step 5.3: Calculate the sector position based on the voltage components in the dq coordinate system, determine the spatial vector action sequence and time allocation;

[0043] Step 5.4: Use the look-up table method to generate the conduction time sequence of each power switch device, and combine the dead-time compensation algorithm to eliminate the inherent delay of the switch devices;

[0044] Step 5.5: Generate complementary drive signals that meet the switching frequency requirements through a PWM generator to ensure that the dead-time of the upper and lower bridge arm switching tubes is consistent;

[0045] Step 5.6: Perform hardware-level filtering on the drive signal to suppress high-frequency harmonic interference;

[0046] Step 5.7: Synchronously transmit the processed drive signal to the six-way power switching devices of the static excitation converter to achieve coordinated control of the generator excitation current and the charge / discharge current of the energy storage unit.

[0047] Preferably, step 1) includes:

[0048] Step 1.1: Use a sliding-mode observer to estimate the rotor flux position and amplitude in real time and construct a synchronous rotating coordinate system;

[0049] Step 1.2: Convert the three-phase current of the control winding into current components in the two-phase stationary αβ coordinate system through Clarke transformation;

[0050] Step 1.3: Based on the rotor flux angle obtained in step 1.1, convert the αβ coordinate system current into the excitation current component and torque current component in the dq rotating coordinate system through Park transformation;

[0051] Step 1.4: Establish a double closed-loop control structure. The outer loop is the excitation current control loop oriented by the rotor flux, and the inner loop is the torque current control loop (objective: maintain the rotor flux constant during steady-state operation to provide a stable basis for power regulation);

[0052] Step 1.5: Convert the decoupled current command into a drive signal through SVPWM technology to control the static excitation converter.

[0053] Preferably, the power distribution table is pre-generated through offline simulation, specifically:

[0054] Step a: Based on the rated parameters of the diesel generator set, the characteristics of the energy storage unit, and typical load scenarios, establish a multi-dimensional simulation model including the generator torque-speed characteristic curve, the charge / discharge efficiency curve of the energy storage unit, and the relationship between the power change rate and system stability;

[0055] Step b: Set boundary conditions in the simulation model: the load power fluctuation range is 0% - 120% of the rated power, the initial value of the energy storage unit SOC is 50%, the charge / discharge power limit is ±80% of the rated power, and the system response time requirement is ≤ 10 ms;

[0056] Step c: Use the genetic algorithm to traverse 2000 groups of candidate parameter combinations and screen out solutions that meet the power regulation error ≤ 3%, the energy storage unit life loss rate ≤ 5% / year, and the reduction of the generator fuel consumption rate ≥ 10%;

[0057] Step d: Store the selected parameter combinations as a three-dimensional table, with dimensions including load power range (low / medium / high), energy storage SOC range (20%-80%), and power change rate levels (slow change / sudden change).

[0058] Step e: Build a simulation platform through MATLAB / Simulink to generate a power distribution table offline.

[0059] The present invention has at least the following beneficial effects:

[0060] 1. Improved dynamic response ability: The present invention realizes the decoupling of excitation and torque current through rotor flux orientation coordinate transformation, combined with real-time power change rate monitoring and a fast response mechanism, significantly shortening the system's response time to sudden load changes, avoiding instantaneous voltage / frequency dips, and improving power supply stability. For example, in step 1 of the present invention, the excitation and torque current are decoupled through rotor flux orientation coordinate transformation, and in step 2, the power change rate is monitored in real time and the fast response mechanism is triggered. The response time is shortened to within 8 ms. When a step load mutation occurs, the voltage fluctuation is ≤±3%, and the recovery time is 6.5 ms, avoiding the instantaneous voltage dip problem caused by offline optimization in the prior art.

[0061] 2. Optimized fuel economy: Based on the piecewise linear power distribution strategy and energy storage cooperative control, the present invention enables the generator to always operate in the high-efficiency load range, reducing fuel consumption. By coordinating the power output of the generator and the energy storage unit through preset rules, fuel waste under low-load conditions is reduced. For example, the piecewise linear power distribution strategy in step 3 of the present invention enables the generator to operate in a high-efficiency range with a fuel efficiency ≥85%. Combining with the energy storage constraint rules, the fuel consumption rate is reduced from 280 g / kW·h to 246 g / kW·h (a decrease of 12%), and the low-load efficiency is improved by 42% compared with the traditional control (≤60%).

[0062] 3. Extended service life of the energy storage unit: By limiting the charge and discharge power and the SOC operating range of the energy storage, overcharging and over-discharging of the battery are avoided, extending the service life of the energy storage unit. The piecewise linear strategy effectively balances the energy storage power compensation and the battery health state, improving the long-term operation reliability of the system.

[0063] Other advantages, objectives, and features of the present invention will be partially reflected by the following description and partially understood by those skilled in the art through the research and practice of the present invention. Brief Description of the Drawings

[0064] Figure 1 It is a schematic flow chart of the intelligent adaptive control method for the diesel generator set described in the present invention. Detailed Embodiments

[0065] The following further elaborates on the present invention in conjunction with embodiments, so that those skilled in the art can implement it with reference to the text of the specification.

[0066] It should be noted that, unless otherwise specified, the experimental methods described in the following implementation schemes are all conventional methods, and the reagents and materials, unless otherwise specified, can be obtained from commercial channels; in the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installation", "connection", and "setting" should be understood in a broad sense. For example, they can be fixedly connected and set, or detachably connected and set, or integrally connected and set. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. The orientation or positional relationship indicated by the terms "lateral", "longitudinal", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention.

[0067] Figure 1 The flow schematic diagram of the intelligent adaptive control method for a diesel generator set of the present invention is shown, which includes the following steps:

[0068] Step 1) Decouple the control winding current of the dual-winding induction generator into the excitation current component and the torque current component through rotor flux-oriented coordinate transformation;

[0069] Step 2) Real-time collect the load power fluctuation signal and calculate the power change rate, and trigger the fast response mechanism when the power change rate exceeds the set threshold;

[0070] Step 3) Based on the piecewise linearized power distribution strategy of preset rules, match the optimal generator output power value and the charge and discharge power value of the energy storage unit in the power distribution table generated offline according to the load power change rate and the current state of charge (SOC) of the energy storage unit;

[0071] Step 4) Adjust the control winding current component through a proportional-integral controller to achieve rapid adjustment of the generator electromagnetic torque;

[0072] Step 5) Use space vector pulse width modulation technology to generate a static excitation converter drive signal to complete the coordinated power output of the generator and the energy storage unit;

[0073] The preset rules include:

[0074] The piecewise linearized rule for dividing the load power range into a low load area, a medium load area, and a high load area,

[0075] A constraint rule that limits the charge and discharge power of the energy storage unit to not exceed 80% of the maximum power and maintains the state of charge between 20% and 80%;

[0076] The power distribution table is pre-generated through off-line simulation and stored in the controller. The controller uses a digital signal processor to realize the real-time call and execution of the power distribution strategy.

[0077] The intelligent adaptive control method of the diesel generator set of the present invention is specifically as follows:

[0078] Step 1) Decouple the control winding current of the dual-winding induction generator into the excitation current component and the torque current component through rotor flux-oriented coordinate transformation; among them, Step 1) includes the following sub-steps:

[0079] Step 1.1: Use a sliding mode observer to estimate the rotor flux position and amplitude in real time and construct a synchronous rotating coordinate system;

[0080] Step 1.2: Convert the three-phase current of the control winding into the current components in the two-phase stationary αβ coordinate system through Clarke transformation;

[0081] Step 1.3: Based on the rotor flux angle obtained in Step 1.1, convert the αβ coordinate system current into the excitation current component and the torque current component in the dq rotating coordinate system through Park transformation;

[0082] Step 1.4: Establish a double closed-loop control structure. The outer loop is the excitation current control loop oriented by the rotor flux, and the inner loop is the torque current control loop (objective: maintain the rotor flux constant during steady-state operation to provide a stable basis for power regulation);

[0083] Step 1.5: Convert the decoupled current command into a drive signal through SVPWM technology to control the static excitation converter.

[0084] Step 2) Collect the load power fluctuation signal in real time and calculate the power change rate. When the power change rate exceeds the set threshold, trigger the fast response mechanism; the said Step 2) includes the following sub-steps:

[0085] Step 2.1: Use current transformers and voltage sensors to synchronously collect the three-phase voltage and current signals of the load;

[0086] Step 2.2: Calculate the real-time values of active power and reactive power based on the instantaneous power theory;

[0087] Step 2.3: Perform a moving average process on the power values for consecutive N sampling periods, and calculate the power change rate ΔP / Δt = [P(k) - P(k - N)] / (N·Ts), where Ts is the sampling period (unit: second, usually Ts = 1 / 10000 second, i.e., 100 μs, to meet the millisecond-level response requirement); where P(k) is the instantaneous power value at the k-th sampling period (unit: kW or MW), P(k - N) is the instantaneous power value at the k - N-th sampling period (the same unit as P(k)), and N is the number of consecutive sampling periods (a positive integer, e.g., N = 10 represents the most recent 10 periods).

[0088] Step 2.4: When the absolute value of ΔP / Δt exceeds the preset power change rate threshold range of ±5% rated power / millisecond, immediately trigger the fast response mechanism; otherwise, enter the conventional control mode.

[0089] The fast response mechanism includes: switching to the high-frequency PID control mode, increasing the control winding current regulation rate to 300% of the rated value, and starting the energy storage unit power compensation program.

[0090] Step 3) Based on the piecewise linearized power distribution strategy according to preset rules, match the optimal generator output power value and the energy storage unit charge / discharge power value in the power distribution table generated offline according to the load power change rate and the current state of charge (SOC) of the energy storage unit; Step 3) includes the following sub-steps:

[0091] Step 3.1: Determine the load interval (low load area, medium load area, or high load area) where the current load power is located according to the load power change rate and the preset piecewise linearized rules.

[0092] Step 3.2: Determine the allowable range of the energy storage unit charge / discharge power according to the current state of charge SOC of the energy storage unit and the preset state of charge constraint rules.

[0093] Step 3.3: In the power distribution table generated offline, match the generator output power distribution coefficient and the energy storage charge / discharge power distribution coefficient corresponding to the current load interval and the allowable range of the energy storage charge / discharge power.

[0094] Step 3.4: Calculate the target output power value of the generator and the target charge / discharge power value of the energy storage unit at the current moment according to the matched distribution coefficients.

[0095] Step 3.5: Verify whether the calculated generator output power and the energy storage charge / discharge power meet the preset power constraint conditions. If not, return to Step 3.1 to re-match.

[0096] Step 3.6: When all constraint conditions are met, output the matched generator output power value and the energy storage charge / discharge power value as the optimal solution.

[0097] Step 4) Adjust the current component of the control winding through a proportional-integral controller to achieve rapid adjustment of the electromagnetic torque of the generator; the said Step 4) includes the following sub-steps:

[0098] Step 4.1: Establish a double closed-loop control structure with the torque control loop as the outer loop and the current control loop as the inner loop (objective: when dynamically responding to load changes, rapidly adjust the power output through the torque loop). For the multi-closed-loop nested design, it includes the outer power loop → the intermediate torque loop → the inner current loop. Step 1.4 is activated during system startup or steady state to ensure the stability of the magnetic flux; Step 4.1 is triggered during load mutation or power regulation demand and takes over the control right; the seamless connection between the two is achieved through a mode switching logic (such as the power change rate threshold); that is, Step 1.4 is the basic field-oriented control, which needs to stabilize the rotor magnetic flux first (the outer-loop excitation current, equivalent to the outer power loop) to provide a stable basis for subsequent torque control. Step 4.1 is the dynamic power regulation, which needs to directly respond to load changes (the outer-loop torque, equivalent to the intermediate torque loop) and rapidly adjust the output through the current loop. Step 1.4 belongs to the underlying current decoupling control, and Step 4.1 belongs to the upper-layer power coordination control. The two belong to different control stages;

[0099] Step 4.2: According to the target output power value of the generator and the target charge-discharge power value of the energy storage unit output in Step 3, calculate the reference value of the current component of the control winding through the power-current mapping relationship;

[0100] Step 4.3: Use a digital PI controller to adjust the current component of the control winding. The PI controller parameters are set as follows: proportional coefficient KP = 2.5, integral coefficient KI = 0.8, and sampling period T = 500 μs;

[0101] When performing closed-loop adjustment on the current component of the control winding, set the current loop bandwidth to 200 Hz, the adjustment period to 1 ms, and the output limit value to 150% of the rated current;

[0102] Step 4.5: When it is detected that the output of the integral term exceeds 80% of the limit value, start the integral separation mechanism to temporarily cut off the integral action to prevent integral saturation;

[0103] Step 4.6: Real-time collect the feedback value of the current component of the control winding through a current sensor, compare it with the reference value calculated in Step 4.2, and generate a current adjustment error signal;

[0104] Step 4.7: Input the current adjustment error signal into the PI controller to output a control winding voltage modulation signal;

[0105] Step 4.8: Convert the voltage modulation signal into drive pulses through SVPWM technology to control the power switching devices of the static excitation converter.

[0106] Step 5) Generate the driving signal of the static excitation converter by using space vector pulse width modulation technology to complete the coordinated power output of the generator and the energy storage unit; the step 5) includes the following sub-steps:

[0107] Step 5.1: Perform Clarke transformation on the voltage modulation signal output in step 4 to convert the voltage components in the three-phase static coordinate system into the two-phase static αβ coordinate system;

[0108] Step 5.2: Perform Park transformation according to the rotor flux orientation angle to convert the voltage components in the αβ coordinate system to the dq rotating coordinate system;

[0109] Step 5.3: Calculate the sector position based on the voltage components in the dq coordinate system to determine the space vector action sequence and time allocation;

[0110] Step 5.4: Use the look-up table method to generate the conduction time sequence of each power switch device, and combine the dead-time compensation algorithm to eliminate the inherent delay of the switch device;

[0111] Step 5.5: Generate complementary driving signals that meet the switching frequency requirements through the PWM generator to ensure that the dead-time of the upper and lower bridge arm switching tubes is consistent;

[0112] Step 5.6: Perform hardware-level filtering on the driving signal to suppress high-frequency harmonic interference;

[0113] Step 5.7: Synchronously transmit the processed driving signal to the six-way power switch devices of the static excitation converter to realize the coordinated control of the generator excitation current and the charge and discharge current of the energy storage unit.

[0114] The preset rules include:

[0115] The piecewise linearization rule that divides the load power range into a low load area, a medium load area, and a high load area,

[0116] The constraint rule that limits the charge and discharge power of the energy storage unit not to exceed 80% of the maximum power and maintains the state of charge at 20% - 80%;

[0117] The power distribution table is pre-generated through offline simulation and stored in the controller. The controller uses a digital signal processor to realize the real-time call and execution of the power distribution strategy; among them, the power distribution table is pre-generated through offline simulation specifically as:

[0118] Step a: Based on the rated parameters of the diesel generator set, the characteristics of the energy storage unit, and typical load scenarios, establish a multi-dimensional simulation model including the generator torque-speed characteristic curve, the charge and discharge efficiency curve of the energy storage unit, and the relationship between the power change rate and system stability;

[0119] Step b: Set boundary conditions in the simulation model: the load power fluctuation range is 0% - 120% of the rated power, the initial value of the energy storage unit SOC is 50%, the charge and discharge power limit is ±80% of the rated power, and the system response time requirement is ≤ 10 ms;

[0120] Step c: Use the genetic algorithm to traverse 2000 groups of candidate parameter combinations, and select solutions that meet the power regulation error ≤ 3%, the energy storage unit life loss rate ≤ 5% / year, and the generator fuel consumption rate reduction ≥ 10%;

[0121] Step d: Store the selected parameter combinations as a three-dimensional table, with dimensions including load power range (low / medium / high), energy storage SOC range (20% - 80%), and power change rate gear (slow change / sudden change);

[0122] Step e: Build a simulation platform through MATLAB / Simulink to generate a power distribution table offline.

[0123] Among them, in the double-winding induction generator system, a sliding-mode observer is used to estimate the position and amplitude of the rotor magnetic flux in real time. The following is the specific algorithm derivation process of the sliding-mode observer: First, the voltage equation of the double-winding induction generator in the stationary αβ coordinate system is: where, u sα and u sβ are the components of the stator voltage in the αβ coordinate system, i sα and i sβ are the components of the stator current in the αβ coordinate system, R s is the stator resistance, ψ sα and ψ sβ are the components of the stator magnetic flux in the αβ coordinate system.

[0124] The relationship between the stator magnetic flux and the rotor magnetic flux is:

[0125] where, L s is the stator self-inductance, L m is the mutual inductance, i rα and i rβ are the components of the rotor current in the αβ coordinate system. The basic idea of the sliding-mode observer is to construct a sliding surface so that the state trajectory of the system can reach and stay on the sliding surface within a finite time.

[0126] Define the sliding surface as:

[0127]

[0128] where, and are the observed stator current components.

[0129] The equation of the sliding mode observer can be expressed as:

[0130]

[0131] Where, is the leakage coefficient, L r is the rotor self-inductance, k s is the sliding mode gain, and sign(·) is the sign function.

[0132] Through the above sliding mode observer equation, the observed rotor flux linkage can be obtained as:

[0133]

[0134] This is the flux linkage calculation method shown in Equation 1-235.

[0135] Adjust the observer gain through the adaptive law

[0136] The selection of the sliding mode gain k s is crucial for the performance of the sliding mode observer. If the gain is too large, it will cause chattering in the system; if the gain is too small, the system may not converge to the sliding mode surface. Therefore, an adaptive law is needed to adjust the observer gain.

[0137] The adaptive law is: Where, γ is the adaptive gain, is the modulus of the sliding mode surface. The working principle of the adaptive law is: when the modulus of the sliding mode surface |s| is large, it means that the system state is far from the sliding mode surface, and at this time, the sliding mode gain k s needs to be increased to accelerate the convergence speed of the system state to the sliding mode surface; when the modulus of the sliding mode surface |s| is small, it means that the system state is already close to the sliding mode surface, and at this time, the sliding mode gain k s can be appropriately reduced to reduce chattering.

[0138] Power distribution table generation process

[0139] When generating the power distribution table, a genetic algorithm is used to screen the optimal parameter combinations. The fitness function of the genetic algorithm is used to evaluate the quality of each parameter combination, and our goal is to simultaneously optimize the power regulation error, the life loss rate of the energy storage unit, and the fuel consumption rate of the generator.

[0140] Let P load be the load power, P g be the generator output power, P s be the charge and discharge power of the energy storage unit, SOC is the state of charge of the energy storage unit, and t is the time. The power regulation error can be expressed as:

[0141] Among them, T is the simulation time. The life loss rate of the energy storage unit can be estimated by the charge-discharge depth and the number of charge-discharge cycles. Assuming that the life loss rate of the energy storage unit is proportional to the square of the charge-discharge depth and proportional to the number of charge-discharge cycles, then the life loss rate E of the energy storage unit s can be expressed as:

[0142]

[0143] where N is the number of charge-discharge cycles, SOC max and SOC min are the maximum and minimum state of charge of the energy storage unit during each charge-discharge process, respectively.

[0144] The fuel consumption rate E of the generator f can be calculated by the torque-speed characteristic curve and the fuel consumption rate curve of the generator. Let T g be the torque of the generator, n g be the speed of the generator, and f(T g , n g ) be the fuel consumption rate function of the generator, then the fuel consumption rate E of the generator f can be expressed as:

[0145]

[0146] Taking the above three indicators into comprehensive consideration, the fitness function of the genetic algorithm can be expressed as:

[0147] F = w1E p + w2E s + w3E f ; where w1, w2, and w3 are weight coefficients, and w1 + w2 + w3 = 1. This is the specific form of the fuel loss objective function of Equation 1-197.

[0148] Specific criteria for parameter screening

[0149] When traversing 2000 groups of candidate parameter combinations using the genetic algorithm, the optimal solution needs to be screened according to the following constraint conditions:

[0150] 1. Power regulation error constraint: E p < 3%, that is, the power regulation error does not exceed 3% of the load power.

[0151] 2. Life loss rate constraint of the energy storage unit: E S ≤ 5% / year, that is, the life loss rate of the energy storage unit does not exceed 5% per year.

[0152] 3. Fuel consumption rate constraint of the generator: E f Reduction ≥ 10%, that is, through optimizing the power distribution, the fuel consumption rate of the generator is reduced by at least 10%.

[0153] Only the parameter combinations that simultaneously satisfy the above three constraint conditions will be stored as valid solutions in the power distribution table. In this way, it can be ensured that the generated power distribution table can achieve high efficiency of power regulation, long life of the energy storage unit, and low fuel consumption of the generator while meeting the system performance requirements.

[0154] Specific implementation example of the present invention:

[0155] In an industrial power supply scenario, a diesel generator set with a rated power of 200 kW (rated voltage 400 V, rated frequency 50 Hz) is configured, and the energy storage unit is a 200 kWh lithium-ion battery pack (maximum charge and discharge power 80 kW, SOC operating range 20% - 80%). The system operates under a step load condition, with an initial load of 100 kW (50% of the rated power), and a sudden load increase of 60 kW (reaching 160 kW, 80% of the rated power) at t = 0, and the power change rate is 60 kW / ms (exceeding the threshold of 5% of the rated power / ms = 10 kW / ms).

[0156] The control process is as follows:

[0157] Real-time signal acquisition and decoupling control:

[0158] The three-phase current of the control winding is converted into αβ coordinate system components through Clarke transformation, and combined with the rotor flux linkage angle estimated by the sliding mode observer (steps 1.1 - 1.3), the d-axis excitation current (0.8 p.u.) and q-axis torque current (1.2 p.u.) are obtained through Park transformation.

[0159] The double closed-loop control structure is automatically activated: the excitation current loop maintains the stability of the rotor flux linkage (step 1.4), laying the foundation for subsequent torque regulation.

[0160] Fast response mechanism is triggered:

[0161] The power sensor detects that the load suddenly increases from 100 kW to 160 kW, and ΔP / Δt = 60 kW / ms is calculated continuously for 10 sampling periods (1 ms) (steps 2.3 - 2.4).

[0162] The system immediately switches to the high-frequency PID control mode, and the current regulation rate is increased to 300% of the rated value (step 2.5), and at the same time, the power compensation program of the energy storage unit is started.

[0163] Piecewise linear power distribution:

[0164] According to the power change rate (60 kW / ms) and the current SOC (50%), the optimal distribution coefficient in the high-load area and under sudden change conditions is matched by looking up the table (steps 3.1 - 3.3).

[0165] It is calculated that the target power of the generator is 140 kW (70% load), and the discharge power of the energy storage unit is 20 kW (step 3.4).

[0166] Verify that the power distribution is within the generator overload capacity (120%) and the energy storage constraint (SOC 20%-80%), and confirm its effectiveness (steps 3.5 - 3.6).

[0167] Fast adjustment of electromagnetic torque:

[0168] The q-axis current reference value of 1.4 p.u. is obtained through the power-current mapping and adjusted by a digital PI controller (KP = 2.5, KI = 0.8) (steps 4.2 - 4.3).

[0169] When it is detected that the integral term is close to the limit value of 80%, the integral action is automatically removed to prevent saturation (step 4.5).

[0170] The current loop bandwidth is 200 Hz, and the actual current is adjusted from 1.2 p.u. to 1.4 p.u. within 1 ms, with an error < 3%.

[0171] Realization of coordinated power output

[0172] After the voltage modulation signal undergoes Clarke-Park transformation, six-way driving pulses are generated through SVPWM (steps 5.1 - 5.3).

[0173] The dead-time compensation algorithm eliminates the switching delay, ensuring that the excitation converter completes power adjustment within 8 ms (steps 5.4 - 5.5).

[0174] Finally, the generator outputs 140 kW, the energy storage unit discharges 20 kW, the system voltage fluctuation is controlled within ±3%, and the recovery time is 6.5 ms, meeting the technical index requirements (step 5.7).

[0175] Implementation effect:

[0176] The response time of this control method is shortened to within 8 ms during load mutation, a 60% improvement compared to the existing technology (20 ms); the generator always operates in the range where the fuel efficiency ≥ 85%, reducing fuel consumption by 12% compared to traditional control; the SOC of the energy storage unit is maintained at 48% - 52%, effectively extending the battery life.

[0177] Example of the sliding mode observer and power distribution table of the present invention:

[0178] 1. System configuration

[0179] Diesel generator set: Rated power 200 kW, rated voltage 400 V, rated frequency 50 Hz, maximum torque 450 N·m (100% load), fuel consumption rate 280 g / kW·h (full load), rated amplitude of rotor magnetic flux 1.2 p.u.

[0180] Energy storage unit: 200 kWh lithium-ion battery pack, single cell voltage 3.2 V, maximum charge and discharge power 80 kW, SOC operating range 20% - 80%

[0181] Control platform: TITMS320F28379D digital signal processor, sampling frequency 10 kHz, supporting SVPWM hardware acceleration

[0182] 2. Observer implementation

[0183] Steps 1.1 - 1.3 sliding mode observer algorithm

[0184] According to step 1.1 of claim 1, a sliding mode observer is used to estimate the rotor magnetic flux in real time:

[0185] Stator voltage equation (stationary αβ coordinate system):

[0186] (Equation 1 - 235 corresponding formula transformation);

[0187] Sliding mode surface definition:

[0188] Observer equation:

[0189] (Equation 1 - 235 core algorithm);

[0190] Adaptive gain adjustment:

[0191] (Realize the magnetic flux stable control of step 1.4).

[0192] 3. Power distribution table generation

[0193] Steps a - e genetic algorithm optimization

[0194] Offline simulation method to generate a power distribution table:

[0195] (1) Fitness function (Equation 1 - 197 extension): F = 0.5E p +0.3E s +0.2E f ;

[0196] Among them: Power regulation error

[0197] Energy storage life loss E s= ∑(SOC max - SOC min )² ≤ 5% / year;

[0198]

[0199] (2) Parameter screening criterion: Generator output power constraint: 0.5P GN ≤ P g ≤ 1.2P GN ; Energy storage charge and discharge constraint |P s | ≤ 0.8P GN ; SOC maintenance range: 20% ≤ SOC ≤ 80%.

[0200] (3) Generated results (partial table data):

[0201]

[0202] 4. Step load response verification experimental conditions:

[0203] Initial load: 100kW (50% rated power);

[0204] Sudden change load: +60kW (to 160kW, 80% rated power);

[0205] Power change rate: 60kW / ms (exceeding the threshold of 5% rated power / ms);

[0206] Control process:

[0207] (1) Fast response trigger (Claim 2, Step 2.5): Switch to high-frequency PID mode, and the current regulation rate is increased to 300% of the rated value; The energy storage unit starts power compensation and outputs 20kW within 5ms;

[0208] (2) Power distribution execution (Claim 3, Step 3.4): Look up the table to match the sudden change condition coefficient in the high load area: Generator 140kW (70% load), Energy storage 20kW;

[0209] Verification: P g = 140kW ≤ 1.2P GN = 240kW, SOC = 50% → 48%;

[0210] (3) Torque regulation result (Claim 4, Step 4.3): PI controller parameters: K p = 2.5, K i = 0.8;

[0211] The current loop bandwidth is 200Hz, and the q-axis current is adjusted from 1.2p.u. to 1.4p.u. within 1ms.

[0212] Experimental results:

[0213] Voltage fluctuation: ±3% (meeting the requirement of ≤5%)

[0214] Recovery time: 6.5 ms (meeting the requirement of ≤8 ms)

[0215] Fuel consumption: 280 g / kW·h → 246 g / kW·h (a 12% decrease).

[0216] Although the embodiments of the present invention have been disclosed above, it is not limited to the applications listed in the specification and embodiments. It can be fully applied to various fields suitable for the present invention. For those familiar with the field, additional modifications can be easily achieved.

Claims

1. An intelligent adaptive control method for diesel generator sets, characterized in that It includes the following steps: Step 1) Decouple the control winding current of the dual-winding induction generator into the excitation current component and the torque current component through rotor flux-oriented coordinate transformation; Step 2) Real-time collect the load power fluctuation signal and calculate the power change rate. When the power change rate exceeds the set threshold, trigger the fast response mechanism; Step 3) Based on the piecewise linear power distribution strategy of preset rules, according to the load power change rate and the current state of charge of the energy storage unit, match the optimal generator output power value and the energy storage unit charge and discharge power value in the power distribution table generated offline; Step 4) Adjust the control winding current component through a proportional-integral controller to achieve fast adjustment of the generator electromagnetic torque; Step 5) Use space vector pulse width modulation technology to generate the driving signal of the static excitation converter to complete the coordinated power output of the generator and the energy storage unit; The preset rules include: The piecewise linearization rule that divides the load power range into a low load area, a medium load area, and a high load area; The constraint rule that limits the charge and discharge power of the energy storage unit not to exceed 80% of the maximum power and the state of charge is maintained between 20% and 80%; The power distribution table is pre-generated through offline simulation and stored in the controller. The controller uses a digital signal processor to realize the real-time call and execution of the power distribution strategy.

2. The intelligent adaptive control method for a diesel generator set according to claim 1, characterized in that Step 2) includes: Step 2.1: Synchronously collect the load three-phase voltage and current signals by using a current transformer and a voltage sensor; Step 2.2: Calculate the real-time values of active power and reactive power based on the instantaneous power theory; Step 2.3: Perform a moving average process on the power values of continuous N sampling periods, and calculate the power change rate ΔP / Δt = [P(k) - P(k - N)] / (N·Ts), where Ts is the sampling period; where, P(k) is the instantaneous power value of the kth sampling period, P(k - N) is the instantaneous power value of the (k - N)th sampling period, and N is the number of consecutive sampling periods; Step 2.4: When the absolute value of ΔP / Δt exceeds the preset power change rate threshold range of ±5% rated power / millisecond, immediately trigger the fast response mechanism. Step 2.5: The fast response mechanism includes: switching to the high-frequency PID control mode, increasing the control winding current adjustment rate to 300% of the rated value, and starting the energy storage unit power compensation program.

3. The intelligent adaptive control method for a diesel generator set according to claim 1, wherein, Step 3) includes: Step 3.1: Determine the load interval where the current load power is located according to the load power change rate and the preset piecewise linearization rule; Step 3.2: Determine the allowable range of the charge and discharge power of the energy storage unit according to the current state of charge SOC of the energy storage unit and the preset state of charge constraint rule; Step 3.3: In the power distribution table generated offline, match the generator output power distribution coefficient and the energy storage charge and discharge power distribution coefficient corresponding to the current load interval and the energy storage charge and discharge power allowable range; Step 3.4: Calculate the target output power value of the generator and the target charge and discharge power value of the energy storage unit at the current moment according to the matched distribution coefficients; Step 3.5: Verify whether the calculated generator output power and energy storage charge-discharge power meet the preset power constraint conditions. If not, return to Step 3.1 to rematch; Step 3.6: When all constraint conditions are met, output the matched generator output power value and energy storage charge-discharge power value as the optimal solution.

4. The intelligent adaptive control method for a diesel generator set according to claim 1, wherein The said Step 4) includes: Step 4.1: Establish a double closed-loop control structure with the torque control loop as the outer loop and the current control loop as the inner loop; Step 4.2: According to the generator target output power value and energy storage unit target charge-discharge power value output in Step 3, calculate the reference value of the control winding current component through the power-current mapping relationship; Step 4.3: Use a digital PI controller to adjust the control winding current component. The PI controller parameters are set as: proportional coefficient KP = 2.5, integral coefficient KI = 0.8, and sampling period T = 500 μs; Step 4.4: When performing closed-loop regulation on the control winding current component, set the current loop bandwidth to 200 Hz, the regulation period to 1 ms, and the output limit value to 150% of the rated current; Step 4.5: When it is detected that the output of the integral term exceeds 80% of the limit value, start the integral separation mechanism to temporarily remove the integral effect and prevent integral saturation; Step 4.6: Real-time collect the feedback value of the control winding current component through a current sensor, compare it with the reference value calculated in Step 4.2, and generate a current regulation error signal; Step 4.7: Input the current regulation error signal into the PI controller to output a control winding voltage modulation signal; Step 4.8: Convert the voltage modulation signal into drive pulses through SVPWM technology to control the power switch devices of the static excitation converter.

5. The intelligent adaptive control method for a diesel generator set according to claim 1, wherein, The said Step 5) includes: Step 5.1: Perform Clarke transformation on the voltage modulation signal output in Step 4 to convert the voltage components in the three-phase stationary coordinate system into voltage components in the two-phase stationary αβ coordinate system; Step 5.2: Perform Park transformation according to the rotor flux orientation angle to convert the voltage components in the αβ coordinate system to the dq rotating coordinate system; Step 5.3: Calculate the sector position based on the voltage components in the dq coordinate system, and determine the spatial vector action sequence and time allocation; Step 5.4: Use the look-up table method to generate the conduction time sequence of each power switch device, and combine the dead-time compensation algorithm to eliminate the inherent delay of the switch device; Step 5.5: Generate complementary drive signals that meet the switching frequency requirements through a PWM generator to ensure that the dead-time of the upper and lower bridge arm switch tubes is consistent; Step 5.6: Perform hardware-level filtering on the drive signals to suppress high-frequency harmonic interference; Step 5.7: Synchronously transmit the processed drive signals to the six-way power switch devices of the static excitation converter to achieve coordinated control of the generator excitation current and the energy storage unit charge-discharge current.

6. The intelligent adaptive control method for a diesel generator set according to claim 1, characterized in that, The said Step 1) includes: Step 1.1: Use a sliding mode observer to estimate the rotor flux position and amplitude in real time and construct a synchronous rotating coordinate system; Step 1.2: Convert the three-phase current of the control winding into current components in the two-phase stationary αβ coordinate system through Clarke transformation; Step 1.3: Based on the rotor flux angle obtained in Step 1.1, transform the αβ coordinate system current into the field current component and torque current component in the dq rotating coordinate system through Park transformation; Step 1.4: Establish a double closed-loop control structure, with the outer loop being the field current control loop oriented by the rotor flux and the inner loop being the torque current control loop; Step 1.5: Convert the decoupled current command into a drive signal through SVPWM technology to control the static excitation converter.

7. The intelligent adaptive control method for a diesel generator set according to claim 1, characterized in that The power distribution table is pre-generated through offline simulation, specifically as follows: Step a: Based on the rated parameters of the diesel generator set, the characteristics of the energy storage unit, and typical load scenarios, establish a multi-dimensional simulation model including the generator torque-speed characteristic curve, the charge-discharge efficiency curve of the energy storage unit, and the relationship between the power change rate and system stability; Step b: Set boundary conditions in the simulation model: the load power fluctuation range is 0% - 120% of the rated power, the initial value of the SOC of the energy storage unit is 50%, the charge-discharge power limit is ±80% of the rated power, and the system response time requirement is ≤10 ms; Step c: Use the genetic algorithm to traverse 2000 groups of candidate parameter combinations, and screen out the solutions that meet the power regulation error ≤ 3%, the energy storage unit life loss rate ≤ 5% / year, and the reduction of the generator fuel consumption rate ≥ 10%; Step d: Store the screened parameter combinations as a three-dimensional table, with dimensions including the load power interval, the energy storage SOC interval, and the power change rate gear; Step e: Build a simulation platform through MATLAB / Simulink to generate the power distribution table offline.

Images