Control method based on flywheel energy storage and super capacitor collaborative optimization

Abstract

The invention relates to the technical field of hybrid energy storage control, and discloses a control method based on flywheel energy storage and super capacitor collaborative optimization, and the method comprises the steps: collecting an energy storage unit state, a power demand signal and a short-term load prediction signal; comparing the signal with a threshold value, and deciding to activate three modes: in a UMC equilibrium response mode, performing optimal power distribution based on a unified marginal cost model; in the economic taxiing standby mode, a pulse taxiing strategy is executed to reduce standby loss; in a feedforward combat readiness standby mode, reversely solving a target standby rotating speed by using the short-term load prediction signal, and adjusting the state of an energy storage unit in advance; and calculating and issuing a control instruction. According to the invention, by introducing the short-term load prediction signal and setting three operation modes, optimization of system operation cost and reduction of standby power consumption are realized, and the active combat readiness capability and response speed of the system for coping with future power impact are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of hybrid energy storage control technology, specifically a control method based on the synergistic optimization of flywheel energy storage and supercapacitors. Background Technology

[0002] Hybrid energy storage systems, which combine flywheel energy storage and supercapacitors, can leverage the energy density advantage of the former and the power density advantage of the latter, and are widely used in applications with severe power fluctuations, such as rail transit and microgrids.

[0003] Existing control strategies typically employ power allocation methods based on filtering algorithms or fixed rules. While these methods can achieve basic power decoupling, they lack a refined consideration of energy storage unit operating losses and aging costs, making it difficult to optimize the overall system operating cost under complex operating conditions.

[0004] In addition, to maintain its energy storage state, the flywheel energy storage unit needs to continuously consume energy to maintain its rotational speed during system idle periods, resulting in high standby power consumption. Meanwhile, traditional control methods are mostly passive responses, only adjusting power after the load changes. When the system faces an impending high-power surge, existing technologies cannot take advantage of the short-term predictability of the load to adjust the energy storage system to its optimal standby state in advance, which limits the improvement of the system's instantaneous response capability and energy utilization efficiency.

[0005] Therefore, this invention proposes a control method based on the coordinated optimization of flywheel energy storage and supercapacitors to address the shortcomings of existing technologies. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a control method based on the coordinated optimization of flywheel energy storage and supercapacitors. This method solves the problem that traditional hybrid energy storage control strategies can only respond passively and fail to achieve coordinated optimization between operational economy, standby power consumption, and active combat readiness response capabilities.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a control method based on the synergistic optimization of flywheel energy storage and supercapacitors, comprising the following steps: Step 1: During the current control cycle, collect the instantaneous rotational speed of the flywheel energy storage unit and the instantaneous terminal voltage of the supercapacitor unit, and receive the power demand signal and the short-term load prediction signal reflecting the future load change trend. Step 2: Compare the power demand signal and the short-term load prediction signal with preset decision thresholds respectively, and determine the target operating mode that the system should activate from the preset operating modes based on the comparison results; the operating modes include at least UMC equalization response mode, economic coasting standby mode and feedforward standby mode. Step 3: Activate the corresponding execution logic according to the determined target operating mode, and calculate the power command of the flywheel energy storage unit and the power command of the supercapacitor unit in the current control cycle. Step four: Convert the power commands of the flywheel energy storage unit and the supercapacitor unit into control signals that can be executed by the underlying power conversion system and send them down. Then return to step one to enter the next control cycle.

[0008] Preferably, the specific process of determining the target operating mode that the system should currently activate in step two includes: The absolute value of the power demand signal is compared with a preset power demand dead zone threshold. The absolute value of the short-term load prediction signal is compared with a preset load prediction impact threshold. When the absolute value of the power demand signal is greater than the power demand dead zone threshold, it is determined that there is a power demand in the system, and the decision is made to activate the UMC equalization response mode. When the absolute value of the power demand signal is less than or equal to the power demand dead zone threshold, and the absolute value of the short-term load prediction signal is less than or equal to the load prediction impact threshold, the system is determined to be in an idle state, and the decision is made to activate the economic coasting standby mode. When the absolute value of the power demand signal is less than or equal to the power demand dead zone threshold, but the absolute value of the short-term load prediction signal is greater than the load prediction impact threshold, it is determined that the system is about to experience a power impact, and the decision is made to activate the feedforward standby mode.

[0009] Preferably, when the UMC balanced response mode is activated, the specific execution process of step three includes: The pre-built unified marginal cost lookup table is invoked, which records the mapping relationship between the operating status of the energy storage unit and the unified marginal cost; Using the power demand signal, instantaneous rotational speed, and instantaneous terminal voltage as inputs, an iterative solution algorithm is used to find a power allocation scheme that meets preset constraints. The preset constraints are: The sum of the output power of the flywheel energy storage unit and the output power of the supercapacitor unit is equal to the power demand signal, and the uniform marginal cost corresponding to the flywheel energy storage unit is equal to the uniform marginal cost corresponding to the supercapacitor unit. The output power of the flywheel energy storage unit and the output power of the supercapacitor unit that meet the preset constraints are respectively used as the power command of the flywheel energy storage unit and the power command of the supercapacitor unit.

[0010] Preferably, when the economic coasting standby mode is activated, the specific execution process of step three includes: Determine whether the instantaneous rotational speed of the flywheel energy storage unit is higher than the lower limit of the preset economic range; If so, the coasting logic is executed, and the power command of the flywheel energy storage unit is set to zero, so that the flywheel energy storage unit decelerates naturally under the effect of standby loss. If not, pulse logic is executed to control the supercapacitor unit to output a preset pulse charging power to the flywheel energy storage unit until the instantaneous speed of the flywheel energy storage unit recovers to the upper limit speed of the preset economic range.

[0011] Preferably, the calibration method for the preset economic range and pulse charging power includes: Based on the unified marginal cost model of the flywheel energy storage unit, the total standby cost, including standby power loss and equivalent aging cost, is calculated. With the goal of minimizing the average overall cost of the complete cycle including the coasting phase and the pulse phase, the upper and lower limits of the speed of the flywheel energy storage unit in standby mode are solved as the preset economic range; The power value corresponding to the operating point with the highest overall system charging efficiency is selected as the pulse charging power.

[0012] Preferably, when the feedforward combat readiness standby mode is activated, the specific execution process of step three includes: Based on the value of the short-term load prediction signal, the unified marginal cost model of the flywheel energy storage unit is used for reverse solution to determine the target standby speed that can respond to future power surges at the expected unified marginal cost. The current instantaneous speed of the flywheel energy storage unit is compared with the target standby speed, and a power command for the flywheel energy storage unit is generated to control the flywheel energy storage unit to accelerate, decelerate or maintain its speed, so that the instantaneous speed approaches the target standby speed.

[0013] Preferably, the specific process of generating power commands for the flywheel energy storage unit to control the acceleration, deceleration, or maintenance of the speed of the flywheel energy storage unit includes: If the instantaneous rotation speed is lower than the target standby rotation speed and the difference between the two exceeds the control dead zone, a negative charging power command is output to increase the instantaneous rotation speed. If the instantaneous speed is higher than the target standby speed and the difference between the two exceeds the control dead zone, a zero power command is output to cause the flywheel energy storage unit to glide and decelerate naturally. If the difference between the current instantaneous speed and the target standby speed is within the control dead zone, a power command is output to offset standby losses in order to maintain the instantaneous speed constant.

[0014] Preferably, when the feedforward combat readiness standby mode is activated, the specific execution process of step three further includes: Monitor the instantaneous terminal voltage of the supercapacitor unit; If the instantaneous terminal voltage is lower than the preset maximum operating voltage, a charging power command is generated for the supercapacitor unit to raise the instantaneous terminal voltage to the maximum operating voltage or a preset voltage value close to the maximum operating voltage, so as to reduce the initial uniform marginal cost of the supercapacitor unit.

[0015] Preferably, the short-term load prediction signal originates from data calculated by the upper-level energy management system based on the scheduling strategy, or from data predicted by an independent load prediction unit based on historical operating data and a time series analysis model.

[0016] Preferably, the specific process of converting the power commands of the flywheel energy storage unit and the supercapacitor unit into control signals executable by the underlying power conversion system in step four includes: The power command of the supercapacitor unit is divided by the instantaneous terminal voltage to obtain the current reference value, and the current reference value is sent to the power conversion system of the supercapacitor unit. The torque reference value is obtained by dividing the power command of the flywheel energy storage unit by the mechanical angular velocity corresponding to the instantaneous rotational speed, and the torque reference value is sent to the motor driver of the flywheel energy storage unit.

[0017] This invention provides a control method based on the synergistic optimization of flywheel energy storage and supercapacitors. It has the following beneficial effects: 1. This invention introduces a short-term load prediction signal and sets a feedforward standby mode. When the control method determines that the system is idle but about to be subjected to a high-power impact, it can use the prediction signal to adjust the state of the flywheel energy storage unit to the optimal target standby speed in advance and prepare the supercapacitor unit to a high-charge state. This avoids the problem that the system will rely too much on the supercapacitor response when the power impact actually arrives due to the flywheel speed being too low, and significantly improves the response speed and stability of the hybrid energy storage system in the face of instantaneous power impact.

[0018] 2. When the system responds to power demand, the present invention adopts the UMC balanced response mode based on unified marginal cost. This mode models the operating loss and standby loss of the flywheel energy storage unit and the aging cost of the supercapacitor unit in a unified manner. By solving the power allocation scheme that makes the marginal costs of the two equal in real time, it ensures that while the system meets the external power demand, its internal comprehensive operating cost (loss and aging) is always kept at the lowest level, thereby improving the economic benefits of the system throughout its entire life cycle.

[0019] 3. This invention addresses the high standby power consumption problem of hybrid energy storage systems under idle conditions by designing an economic coasting standby mode. This mode adopts a pulse coasting strategy, allowing the flywheel energy storage unit to naturally decelerate under standby losses. Only when the speed drops to the lower limit of the economic range is the supercapacitor unit used for efficient pulse charging to restore the speed. This intermittent working mode replaces the traditional standby mode that continuously consumes energy to maintain the speed, effectively reducing the standby power consumption of the system. Attached Figure Description

[0020] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a schematic diagram of the system architecture of the present invention.

[0021] The module includes: 10. Status Acquisition and Preprocessing Module; 20. UMC Modeling and Table Lookup Module; 30. Collaborative Mode Decision Module; 40. Mode Execution Module; 41. UMC Balanced Response Module; 42. Economic Coasting Standby Module; 43. Feedforward Combat Readiness Standby Module; and 50. Command Generation and Issuance Module. Detailed Implementation

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] See attached document Figure 1 With appendix Figure 2 This invention provides a control method based on the coordinated optimization of flywheel energy storage and supercapacitors. The method is executed in a hybrid energy storage system, which may include a flywheel energy storage unit (FES), a supercapacitor unit (SC), a power conversion system (PCS) for energy conversion, a sensing and acquisition unit for monitoring the operating status, and a core controller.

[0024] The core controller is used to execute the method of the present invention and may include multiple functional modules: The status acquisition and preprocessing module 10 is used to acquire the rotational speed of the FES unit, the terminal voltage of the SC unit, the current power demand signal sent by the upper system, and the short-term load prediction signal from the sensing acquisition unit in real time.

[0025] The UMC modeling and lookup module 20 internally stores or embeds the Unified Marginal Cost (UMC) model for FES and SC units. This UMC model can be concretized into a multidimensional lookup table (LUT) that the controller can query at high speed. This lookup table establishes the mapping relationship between the operating status of the energy storage unit (such as voltage, speed, and power) and the unified marginal cost.

[0026] The collaborative mode decision module 30 is used to receive the current power demand signal and the short-term load prediction signal, and decide the operating mode that the system should currently execute according to the preset logic rules, and output the corresponding mode activation command.

[0027] The mode execution module 40 is used to receive mode activation instructions; the mode execution module 40 may include three parallel sub-modules: UMC equalization response module 41, economic coasting standby module 42, and feedforward combat readiness standby module 43; the mode execution module 40 activates one of the three sub-modules according to the received mode activation instructions.

[0028] The instruction generation and distribution module 50 has its input connected to the mode execution module 40 and its output connected to the power conversion system (PCS). The instruction generation and distribution module 50 is used to receive the execution results (such as power reference instructions or state adjustment instructions) from the activated sub-modules and convert them into the underlying control signals of the power conversion system (PCS).

[0029] See attached document Figure 1 This invention provides a control method based on the synergistic optimization of flywheel energy storage and supercapacitors, which may include the following steps: S100, System initialization; The core controller loads the calibration model parameters of the FES unit and SC unit, as well as the multidimensional lookup table data required by the UMC modeling and lookup table module 20; S110, Real-time Status Acquisition; The status acquisition and preprocessing module 10 acquires the rotational speed of the FES unit, the terminal voltage of the SC unit, the current power demand signal, and the short-term load prediction signal in each control cycle. S120, Operation mode decision; The collaborative mode decision module 30 receives the current power demand signal and short-term load prediction signal collected in step S110, and makes a judgment according to the preset logic rules, and outputs a mode activation command. S130, branch mode execution; mode execution module 40 activates one of UMC equalization response module 41, economic coasting standby module 42 or feedforward combat standby module 43 according to the mode activation instruction output in step S120, so as to execute the corresponding control logic. S140, Control command generation and issuance; The command generation and issuance module 50 receives the execution result from the activated sub-module in step S130, calculates it into a control signal for the power conversion system (PCS), and issues it to the PCS for execution. S150, execute repeatedly; the method returns to step S110 and enters the next control cycle.

[0030] To further clarify the implementation of each technical aspect of the present invention, the following will provide a detailed description of the implementation of each functional module involved above and its internal processing flow.

[0031] This step is the preparation phase for method execution, which mainly involves loading the basic model and parameters required for the core controller to run.

[0032] The core controller first loads the predetermined parameters of each hardware unit in the hybrid energy storage system; these parameters may include: the mechanical parameters and motor parameters of the flywheel energy storage unit (FES); and the equivalent series resistance of the supercapacitor unit (SC). The nominal capacitor and the efficiency characteristic curve of the power conversion system (PCS) are known technologies in this field and will not be described in detail here.

[0033] The key to this step is the construction and loading of the Unified Marginal Cost (UMC) model, which is completed or loaded by the UMC modeling and table lookup module 20 during system initialization.

[0034] The goal of the UMC model is to quantify the total cost incurred by a system at a specific operating point for outputting unit power. Defined as instantaneous power loss Equivalent aging cost sum: ; Corresponding uniform marginal cost Defined as total cost For output power Partial derivatives: ; This model was applied to both SC and FES elements.

[0035] S101, Constructing the UMC model of the supercapacitor unit ( Instantaneous power loss of SC unit Mainly its equivalent series resistance Joule heat loss on top: ; in, The output power of the SC unit. This represents the real-time terminal voltage of the SC cell. This represents the real-time current of the SC cell. This is the equivalent series resistance of the SC unit.

[0036] Equivalent aging cost of SC unit The aging rate, accelerated by operating conditions (especially high voltage and high current), is quantified using weighting factors. Converted to equivalent power: ; in, This is the aging cost reduction factor for SC units. This is the aging characteristic function of the SC cell, which reflects the nonlinear influence of voltage and current on the lifetime loss of the SC cell.

[0037] Therefore, the uniform marginal cost of the SC unit The calculation is as follows: ; S102, Construct the UMC model of the flywheel energy storage unit ( Instantaneous power loss of the FES unit Including standby power consumption (such as bearing friction, wind resistance) and operating losses (e.g., motor copper losses, iron losses): ; in, The output power of the FES unit, The rotational speed of the FES unit. The output torque of the FES unit, Rotational speed The corresponding mechanical angular velocity, Rotational speed The function, Rotational speed and torque The function.

[0038] Equivalent aging cost of FES unit This is used to quantify the aging caused by changes in rotational speed (affecting bearing wear) and power (affecting mechanical stress), through weighting coefficients. Conversion: ; in, This is the aging cost discount factor for the FES unit. This is the aging characteristic function of the FES unit, which reflects the impact of speed and power change rate on the life loss of the FES unit.

[0039] Therefore, the uniform marginal cost of the FES unit The calculation is as follows ( As standby power consumption, (The partial derivative is 0) ; S103, implements the engineering loading of the UMC model; given the formula and formula The analytical expression is highly nonlinear and depends on multiple real-time states. This step is implemented using offline calibration and online lookup table (LUT) methods.

[0040] Offline phase: Through simulation calculations or experimental bench calibration, in the state space of the SC unit ( , ) and the state space of the FES unit ( Multidimensional scanning was performed on the surface.

[0041] At each discrete state point of the scan, according to the formula and formula Calculate the corresponding and The precise value.

[0042] All discrete state points and their corresponding The calculation results are stored in the non-volatile memory of the core controller, forming a multidimensional lookup table: ; ; Online Phase: The loading in step S100 refers to the UMC modeling and lookup module 20 reading these LUT data from non-volatile memory into the cache; subsequent steps (such as S130) only need to be executed according to the real-time status, by querying the LUT and combining it with a multi-dimensional interpolation algorithm (such as bilinear interpolation or trilinear interpolation), to instantly obtain the data. This avoids the complex online calculation of partial derivatives.

[0043] This step is the data input phase of the method operation, which is performed by the status acquisition and preprocessing module 10 in each control cycle. This process is executed within the timeframe, providing a real-time data foundation for subsequent decision-making (S120) and execution (S130) steps.

[0044] S111, collects the internal operating status of the energy storage unit; the status acquisition and preprocessing module 10 obtains the instantaneous rotational speed of the FES unit via the sensing acquisition unit. ;Should The signal can originate from an encoder, rotary transformer, or Hall effect sensor mounted on the rotor shaft of the FES unit; the status acquisition and preprocessing module 10 simultaneously acquires the instantaneous terminal voltage of the SC unit via a high-precision voltage sampling circuit (part of the sensing acquisition unit). .

[0045] For the collected and The original signal (if it is an analog signal) is converted into a digital signal by the analog-to-digital converter (ADC) inside the state acquisition and preprocessing module 10. The state acquisition and preprocessing module 10 can further perform preprocessing on the converted digital signal, such as applying digital filtering algorithms (e.g., low-pass filtering or Kalman filtering) to eliminate high-frequency noise and ensure the stability of subsequent calculations.

[0046] S112 receives external commands and predictive signals; the status acquisition and preprocessing module 10 receives the current power demand signal. ;Should The signal can be calculated by an upper-level energy management system (EMS) according to the overall system scheduling strategy and sent to the core controller via a communication bus (e.g., CAN bus or Ethernet); in another embodiment, It can also be measured and calculated in real time by a local load sensor.

[0047] The status acquisition and preprocessing module 10 simultaneously receives short-term load prediction signals. ;Should The signal can be generated by a separate load forecasting unit that operates based on historical data and a specific forecasting model (such as time series analysis or machine learning models), using the trend of power demand changes over a short future time window (e.g., seconds to minutes) as the basis for prediction. The signal is sent to the status acquisition and preprocessing module 10 via the communication bus.

[0048] The status acquisition and preprocessing module 10 will process all the valid data acquired in S111 and S112 ( , , , Update the internal registers or shared data area of ​​the core controller so that the cooperative mode decision module 30 in step S120 and the mode execution module 40 in step S130 can call it in the current control cycle.

[0049] This step is executed by the collaborative mode decision module 30, which is the top-level logic scheduling core of the method and is used to decide the operating mode that the system should enter based on the real-time signals collected in step S110.

[0050] S121, Set decision thresholds; To achieve logical judgment of signals, at least two threshold parameters are preset or loaded in the cooperative mode decision module 30: Power demand dead zone threshold : A preset small positive value; when absolute value At that time, determine the current power demand signal. Approximately zero; conversely, if Then judge Not zero.

[0051] Load prediction threshold A preset positive value used to define Does the signal constitute a significant impact? When absolute value At that time, determine the short-term load prediction signal Approximately zero (or no impact); conversely, if Then judge Significantly non-zero.

[0052] S122, execution mode switching logic; the collaborative mode decision module 30 acquires the data collected in S110. and Following the signal, within the current control cycle, based on the threshold defined in S121, the following preset, priority-based logic rules are executed to output a unique mode activation command: Logical Rule 1: Collaborative Mode Decision Module 30 first makes a judgment. ;like Then it is determined that the system currently has a clear power response requirement (i.e. (not zero); at this time, regardless of Regardless of the signal value, the decision to activate [Mode 1] will be made immediately, and an instruction to activate the UMC equalization response module 41 will be output.

[0053] Logical Rule 2: If Logical Rule 1 is not triggered (i.e.) ,judge (Approximately zero), the collaborative mode decision module 30 will continue to make judgments. If at this time (Right now If the value is approximately zero, then the system is determined to be in an idle standby state; at this time, the decision is made to activate [Mode 2] and output the instruction to activate the economic coasting standby module 42.

[0054] Logical Rule 3: If Logical Rule 1 is not triggered (i.e.) ),but (Right now If the value is significantly non-zero, it indicates that although the system currently has no power demand, it anticipates an impending power surge. In this case, the decision is made to activate [Mode 3], and an instruction to activate the feedforward standby module 43 is output.

[0055] The collaborative mode decision module 30 sends the unique mode activation instruction generated in S122 to the mode execution module 40 in step S130.

[0056] This step is the core execution phase of the method, where the pattern execution module 40 activates one of the three parallel sub-modules according to the instructions in step S120.

[0057] Detailed implementation of [Mode 1]: UMC equalization response; when the cooperative mode decision module 30 (step S120) judges the current power demand signal Not zero (i.e.) When [mode 1] is reached, the mode activation command for [mode 1] is output.

[0058] After receiving the instruction, the mode execution module 40 activates the UMC equalization response module 41; the UMC equalization response module 41, within the current control cycle, uses the data collected in step S110... , , And the UMC lookup table loaded in step S100 ( and The input is a set of power commands that satisfy specific constraints. and .

[0059] S131, Set constraints; The power command solved by the UMC equalization response module 41 must simultaneously satisfy the following two constraints: Power balance constraints: ; UMC equilibrium constraint: ; in, and The output power of the SC and FES elements to be solved; and This represents the uniform marginal cost of the two units at their respective operating points, obtained from the UMC lookup table.

[0060] S132, execute the equilibrium solution algorithm; to ensure the real-time control cycle ( Within the UMC equalization response module 41, the constraint equations in S131 are solved quickly, and an iterative solution process is performed. This process can utilize power balance constraints to simplify the problem to a single variable (e.g., The root-finding problem, because can be Export.

[0061] This solution algorithm can be specifically implemented as a bisection method or a similar numerical iterative algorithm, with the following steps: S132a, Initialize search range; Set upper bound of search and the lower realm This range can be set as the physical maximum / minimum output power limit of the SC unit; S132b, Select the test power; in one iteration, select the test power point. ; S132c, calculate the associated state; calculate the test power point of the FES cell according to the power balance constraint formula. Simultaneously, the power change rate of the FES unit is calculated. ,in This refers to the FES unit power command issued in the previous control cycle (step S140). To control cycle duration; S132d, query the UMC value; using the lookup table loaded in step S100, combined with the real-time status collected in step S110. and Query the UMC values ​​of the two units at the test point: ; ; S132e, calculate the equilibrium error and update the interval; define the equilibrium error. ; like This indicates that the marginal cost of the SC unit is high and should be reduced. The allocation is updated, and the upper bound of the search is updated: ; like This indicates that the marginal cost of the FES unit is high and should be increased. The allocation is updated, and the search lower bound is updated: .

[0062] S132f, determine convergence; repeat S132b to S132e until... The absolute value is less than the preset convergence precision. Or, until the preset maximum number of iterations is reached.

[0063] S133, generate power command; after the iteration of S132f converges, the UMC equalization response module 41 will use the power command obtained from the last iteration. and As the final power command of the current control cycle, and .

[0064] S134, Output command; UMC equalization response module 41 will obtain the solution... and The output is sent to the instruction generation and distribution module 50 in step S140.

[0065] Detailed implementation of [Mode 2]: Economic coasting standby; when the cooperative mode decision module 30 (step S120) judges the current power demand signal Approximately zero (i.e.) ), and short-term load prediction signals It is also approximately zero (i.e.) When ), output the mode activation command for [Mode 2].

[0066] After receiving the instruction, the mode execution module 40 activates the economic coasting standby module 42; the economic coasting standby module 42 executes a pulse coasting strategy to minimize the overall standby cost of the system in the idle state; the execution of this strategy depends on the parameters pre-calibrated in step S100.

[0067] S135, calibrate the economic range and pulse power; in the initialization phase of step S100, the total cost in standby mode has been calibrated according to the UMC model of the FES unit. .

[0068] Using the total standby cost formula Calculate the standby cost of the FES unit: ; in, This refers to the standby losses of the FES unit (such as wind resistance and bearing friction). For the FES unit at a constant speed (i.e. The equivalent aging cost under ( ).

[0069] Based on this The model, and the PCS conversion efficiency characteristics of the SC unit charging the FES unit, were calculated offline and an economic range was determined. and a pulse charging power .

[0070] Economic range These are the upper and lower limits of the rotational speed of the FES unit in standby mode, and their selection is aimed at minimizing the average overall cost of a complete coasting pulse cycle; pulse charging power. The power value corresponding to the highest total system efficiency (combining the PCS converter efficiency of the SC unit and the PCS converter and motor efficiency of the FES unit) when the SC unit is calibrated to charge the FES unit (i.e., the FES unit absorbs power).

[0071] S136, execute the glide phase logic; after the economic glide standby module 42 is activated, in each control cycle, it first acquires the FES unit speed collected in step S110. ;judge Is it greater than the lower limit of the economic range? ;like The economic coasting standby module 42 will execute the coasting logic: output power command. and .

[0072] This zero-power command causes the PCS converter of the FES unit to stop working, and the FES unit rotor... (Standby power loss) naturally reduces speed, thus decreasing its rotational speed. Slow descent.

[0073] S137 executes the Pulse phase logic; the economic coasting standby module 42 continuously monitors... If during the taxiing phase of S136, it is detected that... If this occurs, the pulse logic is immediately triggered: the power command of the output FES unit is executed. (The negative sign indicates that the FES unit absorbs power); output the power command of the SC unit. ; in The SC unit provides The actual output power required is equal to In addition to the SC unit PCS converter, FES unit PCS converter and motor Total loss corresponding to the operating point.

[0074] S138, termination and loop of the pulse phase; during the pulse logic execution of S137, the economic coasting standby module 42 continues to monitor in subsequent control cycles. ;like This indicates that the FES unit speed has recovered to the upper limit of the economic range, and the pulse logic terminates; the economic coasting standby module 42 immediately returns to the coasting stage logic of S136 and outputs again. and The FES unit then resumed its natural gliding motion.

[0075] S139, Output command; Power command generated by the economy coasting standby module 42 in S136 or S137 ( and The instruction is output to the instruction generation and distribution module 50 in step S140.

[0076] Detailed implementation of [Mode 3]: Feedforward standby; when the cooperative mode decision module 30 (step S120) judges the current power demand signal Approximately zero (i.e.) However, short-term load prediction signals Significantly non-zero (i.e.) When ), output the mode activation command for [Mode 3].

[0077] After receiving the instruction, the mode execution module 40 activates the feedforward standby module 43; the feedforward standby module 43 utilizes... The signal immediately interrupts the economic coasting strategy of [Mode 2] (if activated) and instead executes an active state preparation strategy to adjust the operating state of the FES unit and SC unit to the optimal response start point before the power surge actually arrives.

[0078] S140, calculate the combat readiness rotation speed of the FES unit. The feedforward standby module 43 obtains the short-term load prediction signal from step S110. The feedforward standby module 43 is based on The value is used to solve the UMC model of the FES unit in reverse to determine the optimal standby speed. .

[0079] This solution process can be specifically implemented as follows: Determining the marginal cost of a combat readiness target ;Should The system is responding The expected uniform marginal cost when experiencing power surges of this magnitude; It can be a basis Size dynamic query The resulting variable is either a fixed parameter that represents the efficient response range of the system.

[0080] Determine the combat readiness target power of the FES unit. This value is based on Expected power sharing ratio of FES unit ( The calculation yielded the following: .

[0081] The feedforward standby module 43 uses the FES unit UMC lookup table loaded in the reverse lookup or numerical solution step S100. ), find a rotational speed To satisfy: ; in, For the estimated rate of change of power (e.g. , (This is the preset response time).

[0082] Solving point 3 of S140 yields the following results. That is, it is set to combat readiness speed. .

[0083] S141, execute the state preparation logic of the FES unit; the feedforward standby module 43 obtains the current rotational speed of the FES unit from step S110. and the result calculated in S140 Compare them.

[0084] like ( (Based on the preset speed control dead zone), the system determines that the FES unit speed is insufficient; the feedforward standby module 43 immediately outputs the power command for the FES unit. ,in The preset charging power value used to quickly increase the speed (e.g.) ).

[0085] like The FES unit rotation speed is too high; the feedforward standby module 43 outputs... The command stops the PCS converter of the FES unit, allowing the FES unit to coast and decelerate naturally under standby power loss.

[0086] like The FES unit speed has been determined to be within the specified range; the feedforward standby module 43 outputs... ,in For FES units in Standby loss at high speed; this command is used to compensate for standby loss, ensuring the FES unit speed is precisely maintained. .

[0087] S142 executes the state preparation logic of the SC unit; simultaneously, the feedforward standby module 43 executes the state preparation of the SC unit, with the goal of setting its terminal voltage. Increase to maximum operating voltage (or a close one) combat readiness voltage To maximize its energy storage and its initial marginal cost Reduced to the lowest level.

[0088] The feedforward standby module 43 obtains the current status from step S110. ; like ( (For the preset voltage control dead zone), the feedforward standby module 43 will calculate the charging power command of the SC unit. .

[0089] Should It can be provided by a PCS converter connected to the power grid (if present), or under specific operating conditions (e.g., in S141). and Much higher (At that time) is provided by the FES unit braking.

[0090] S143 outputs combined power commands; the final output of the feedforward standby module 43 is... and ; Determined by the logic of S141; Two requirements must be met: First, the power required by the FES unit in S141 (Right now (corresponding SC unit output power); Second, the charging requirements of the SC unit itself in S142. ( (This is the power absorbed by the SC unit, hence a negative value).

[0091] ; in Typically supplied by the power grid, and not included in the calculation. ,but .

[0092] In one specific embodiment, the charging power required by the FES unit in S141 ( (Provided by the SC unit, therefore) Set as ( The output power of the SC unit after PCS conversion); simultaneously, the charging of the SC unit in S142 (if needed) is independently performed by the grid PCS converter; the feedforward standby module 43 will (from S141) and (From S143) Output to the instruction generation and distribution module 50 of step S140.

[0093] This step is the control output phase of the method, executed by the instruction generation and distribution module 50. This module 50 receives the execution result, i.e., the SC unit power instruction, output from the mode execution module 40 (specifically, its activated sub-module) in step S130. and FES unit power command .

[0094] The function of the instruction generation and distribution module 50 is to convert these two high-level power instructions into specific control signals that can be executed by the underlying power conversion system (PCS).

[0095] S141, calculate the control signals of the SC unit; instruction generation and issuance module 50 acquires the real-time terminal voltage of the SC unit collected in step S110. The instruction generation and distribution module 50 uses the SC unit power instruction. and terminal voltage Calculate the current reference value required for the SC unit. .

[0096] The formula for calculating the current reference value of the SC unit is as follows: ; in, This is the current reference value for the SC unit; This is the power command for the SC unit; This represents the real-time terminal voltage of the SC cell.

[0097] This current reference value The current is then input to the underlying current control loop of the power conversion system (PCS) of the SC unit; the underlying current control loop (e.g., a PI controller) according to... The final pulse width modulation (PWM) signal is calculated from the actual sampled SC cell current. This PWM signal is used to drive the power semiconductor devices in the SC cell PCS converter (e.g., a bidirectional DC / DC converter). The implementation of the underlying current control loop is a well-known technology in the field of power electronic converter control, and will not be described in detail here.

[0098] S142, calculate the control signal of the FES unit; instruction generation and issuance module 50 obtains the real-time rotational speed of the FES unit collected in step S110. The instruction generation and distribution module 50 uses the FES unit power instruction. and rotational speed Calculate the reference torque value required for the FES unit motor. .

[0099] The FES unit torque reference value calculation formula is used as follows: ; in, This is the torque reference value for the FES unit motor; Power command for the FES unit; For FES unit rotation speed The corresponding mechanical angular velocity.

[0100] This torque reference value The input is then fed into the underlying torque or current control loop of the power conversion system (i.e., the motor driver) of the FES unit; the motor driver (e.g., using vector control or direct torque control) according to... Based on the real-time status of the motor (such as rotor position and phase current), the three-phase voltage or PWM signal driving the FES unit motor is calculated; the control algorithm of the motor driver is a well-known technology in the field of motor control, and will not be elaborated here.

[0101] S143, send control signals; the instruction generation and sending module 50 sends the PWM signal of the SC unit PCS converter calculated in S141 and the PWM signal of the FES unit motor driver (or its previous stage control signal, such as the current reference value) calculated in S142 to the corresponding power conversion system (PCS) hardware to drive the FES unit and SC unit to operate according to the instruction power.

[0102] This step is a closed-loop step that ensures the continuity and real-time performance of the control method, guaranteeing its continuous operation.

[0103] S151, in step S140, the instruction generation and issuance module 50 completes the current control cycle (defined as the [number]th cycle). After a control signal (per cycle) is issued, the execution flow of the core controller is reset; this reset operation can be interrupted by a periodic timer (whose cycle is the control cycle). The task may be triggered by the task scheduler, or the task scheduler may schedule the next cycle's task immediately after the current cycle's task has been completed.

[0104] S152, the execution process unconditionally returns to step S110 (real-time status acquisition), marking the... The end of the cycle and the first The start of a control cycle.

[0105] S153, in the At the start of each control cycle, the state acquisition and preprocessing module 10 in step S110 will acquire the instantaneous state values ​​within the new cycle, including... , , and These values ​​reflect the system's performance in the first quarter. The latest state under the control of each cycle, and the latest changes in external instructions.

[0106] S154, the core controller subsequently used these in The data collected periodically is processed sequentially through steps S120 (running mode decision), S130 (branch mode execution), and S140 (control command generation and issuance).

[0107] The control method and system of the present invention can be applied to applications with high requirements for power smoothing, energy recovery and instantaneous power response, such as, but not limited to: rail transit, port gantry cranes, construction machinery, power grid frequency regulation, wind power / photovoltaic grid connection and microgrids.

[0108] The following describes the working process of the on-board hybrid energy storage system (HESS) of the present invention applied to urban rail transit vehicles (such as subways and light rails) in detail.

[0109] In this application scenario, the FES unit and SC unit are installed together on the vehicle to absorb regenerative braking energy and assist traction.

[0110] Current power demand signal The traction control unit (TCU) of the vehicle calculates and issues the command in real time based on the position of the driver's traction / brake handle or the instructions of the Automatic Train Operation (ATO) system. This indicates the traction condition (HESS requires discharge). This indicates regenerative braking mode (HESS requires charging).

[0111] Short-term load prediction signal Provided by the vehicle's ATO system or a location-based driver assistance system; the system can anticipate upcoming high-power demand events (such as acceleration for exiting a station or going uphill) several seconds to tens of seconds in advance based on the vehicle's current location (e.g., obtained via GPS, transponder, or odometer), route map (including gradient information), and forward signals (e.g., knowing that an exit is imminent).

[0112] The specific execution process of this control method in a typical operating cycle of a rail transit vehicle (departure, acceleration, cruising, braking, and entry into the station) is as follows: 1. Vehicle departure acceleration (corresponding to [Mode 1]: UMC equalization response); when the train departs from the station, the TCU issues a large positive power command ( and ); The collaborative mode decision module 30 in step S120 immediately makes a judgment. If the value is not zero, activate [Mode 1]; the UMC equalization response module 41 in step S130 is activated; the UMC equalization response module 41 is activated according to the current... and S110 collected and The iterative solution algorithm in S132 is executed in real time; this algorithm is performed in each control cycle. Within, all are calculated to make Established power allocation command and .

[0113] At this stage, high power density SC cells (typically) High, The lower energy density FES unit responds to most of the instantaneous power demand, while the high energy density FES unit provides a smooth and continuous power output; the UMC model ensures that the two discharge together in a way that minimizes the total cost (loss + aging) throughout the acceleration process.

[0114] 2. Vehicle Cruise / Idle (corresponding to [Mode 2]: Economic Coasting Standby); When the train reaches cruising speed and travels on a straight track, or when it stops at a station for an extended period of time, the TCU command is issued. Approximately zero ( ); Meanwhile, the ATO system determines that there are no immediate high-power events ahead (e.g., the next station is far away, or the stop time is long). It is also approximately zero. ).

[0115] Step S120 activates [Mode 2].

[0116] In step S130, the economic coasting standby module 42 is activated, and the pulse coasting strategy of S135 to S139 is executed.

[0117] The PCS converter of the FES unit is turned off (S136). Flywheel speed from It will naturally decrease due to standby power consumption to minimize standby power consumption.

[0118] when Descending to At time (S137), the SC unit operates at maximum efficiency. The power (S135 calibration) pulses charge the FES unit, causing its rotational speed to quickly recover. ; This mode significantly reduces the average standby power consumption of the FES unit under idle conditions.

[0119] 3. Preparing to leave the station before entering (corresponding to [Mode 3]: Feedforward standby); When the train stops at the station, TCU commands are issued. Approximately zero ( At this point, the ATO system learns that the train will depart in (for example) 5 seconds, and based on this, the ATO system generates a significantly non-zero value. Signal( This signal represents the upcoming acceleration power demand.

[0120] The collaborative mode decision module 30 (S122, logical rule three) in step S120 makes a judgment. but Activate [Mode 3] immediately.

[0121] In step S130, the feedforward standby module 43 is activated.

[0122] The feedforward standby module 43 immediately interrupts any [Mode 2] (coasting) that may be in progress.

[0123] Feedforward standby module 43 according to The value is used to perform the S140 calculation, and the UMC model is solved in reverse to obtain an optimal combat readiness rotation speed. .

[0124] The feedforward standby module 43 executes the logic of S141, if the current... Then immediately instruct the SC unit (S143) to... The power is used to charge the FES unit, causing its rotation speed to increase rapidly. And maintain.

[0125] Simultaneously, the feedforward standby module 43 executes the logic of S142, and (if possible) instructs the SC unit to brake via the overhead contact line or FES (if...). (Too high) Replenish charge to nearby.

[0126] 4. Vehicle regenerative braking (corresponding to [Mode 1]: UMC equalization response); when the train brakes upon entering the station, the TCU issues a large negative power command ( and ).

[0127] Step S120 reactivates [Mode 1].

[0128] UMC equalization response module 41 is activated; at this time When the value is negative, the power balance constraint of S131 becomes ; The algorithm for solving S132 (at this time) and All values ​​are negative, representing absorbed power. Similarly, search for... The optimal solution is to allocate regenerative braking energy to the SC unit and FES unit at the lowest overall cost (considering charging losses and aging).

[0129] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A control method based on the synergistic optimization of flywheel energy storage and supercapacitors, characterized in that, Includes the following steps: Step 1: During the current control cycle, collect the instantaneous rotational speed of the flywheel energy storage unit and the instantaneous terminal voltage of the supercapacitor unit, and receive the power demand signal and the short-term load prediction signal reflecting the future load change trend. Step 2: Compare the power demand signal and the short-term load prediction signal with preset decision thresholds respectively, and determine the target operating mode that the system should activate from the preset operating modes based on the comparison results; the operating modes include at least UMC equalization response mode, economic coasting standby mode and feedforward standby mode. Step 3: Activate the corresponding execution logic according to the determined target operating mode, and calculate the power command of the flywheel energy storage unit and the power command of the supercapacitor unit in the current control cycle. Step four: Convert the power commands of the flywheel energy storage unit and the supercapacitor unit into control signals that can be executed by the underlying power conversion system and send them down. Then return to step one to enter the next control cycle.

2. The control method based on the synergistic optimization of flywheel energy storage and supercapacitor as described in claim 1, characterized in that, The specific process for determining the target operating mode that the system should currently activate in step two includes: The absolute value of the power demand signal is compared with a preset power demand dead zone threshold. The absolute value of the short-term load prediction signal is compared with a preset load prediction impact threshold. When the absolute value of the power demand signal is greater than the power demand dead zone threshold, it is determined that there is a power demand in the system, and the decision is made to activate the UMC equalization response mode. When the absolute value of the power demand signal is less than or equal to the power demand dead zone threshold, and the absolute value of the short-term load prediction signal is less than or equal to the load prediction impact threshold, the system is determined to be in an idle state, and the decision is made to activate the economic coasting standby mode. When the absolute value of the power demand signal is less than or equal to the power demand dead zone threshold, but the absolute value of the short-term load prediction signal is greater than the load prediction impact threshold, it is determined that the system is about to experience a power impact, and the decision is made to activate the feedforward standby mode.

3. The control method based on the synergistic optimization of flywheel energy storage and supercapacitor as described in claim 2, characterized in that, When the UMC balanced response mode is activated, the specific execution process of step three includes: The pre-built unified marginal cost lookup table is invoked, which records the mapping relationship between the operating status of the energy storage unit and the unified marginal cost; Using the power demand signal, instantaneous rotational speed, and instantaneous terminal voltage as inputs, an iterative solution algorithm is used to find a power allocation scheme that meets preset constraints. The preset constraints are: The sum of the output power of the flywheel energy storage unit and the output power of the supercapacitor unit is equal to the power demand signal, and the uniform marginal cost corresponding to the flywheel energy storage unit is equal to the uniform marginal cost corresponding to the supercapacitor unit. The output power of the flywheel energy storage unit and the output power of the supercapacitor unit that meet the preset constraints are respectively used as the power command of the flywheel energy storage unit and the power command of the supercapacitor unit.

4. The control method based on the coordinated optimization of flywheel energy storage and supercapacitor as described in claim 2, characterized in that, When the economic coasting standby mode is activated, the specific execution process of step three includes: Determine whether the instantaneous rotational speed of the flywheel energy storage unit is higher than the lower limit of the preset economic range; If so, the coasting logic is executed, and the power command of the flywheel energy storage unit is set to zero, so that the flywheel energy storage unit decelerates naturally under the effect of standby loss. If not, pulse logic is executed to control the supercapacitor unit to output a preset pulse charging power to the flywheel energy storage unit until the instantaneous speed of the flywheel energy storage unit recovers to the upper limit speed of the preset economic range.

5. The control method based on the coordinated optimization of flywheel energy storage and supercapacitor as described in claim 4, characterized in that, The calibration method for the preset economic range and pulse charging power includes: Based on the unified marginal cost model of the flywheel energy storage unit, the total standby cost, including standby power loss and equivalent aging cost, is calculated. With the goal of minimizing the average overall cost of the complete cycle including the coasting phase and the pulse phase, the upper and lower limits of the speed of the flywheel energy storage unit in standby mode are solved as the preset economic range; The power value corresponding to the operating point with the highest overall system charging efficiency is selected as the pulse charging power.

6. The control method based on the synergistic optimization of flywheel energy storage and supercapacitor as described in claim 2, characterized in that, When the feedforward combat readiness standby mode is activated, the specific execution process of step three includes: Based on the value of the short-term load prediction signal, the unified marginal cost model of the flywheel energy storage unit is used for reverse solution to determine the target standby speed that can respond to future power surges at the expected unified marginal cost. The current instantaneous speed of the flywheel energy storage unit is compared with the target standby speed, and a power command for the flywheel energy storage unit is generated to control the flywheel energy storage unit to accelerate, decelerate or maintain its speed, so that the instantaneous speed approaches the target standby speed.

7. The control method based on the coordinated optimization of flywheel energy storage and supercapacitor as described in claim 6, characterized in that, The specific process of generating power commands for the flywheel energy storage unit to control its acceleration, deceleration, or maintenance of rotational speed includes: If the instantaneous rotation speed is lower than the target standby rotation speed and the difference between the two exceeds the control dead zone, a negative charging power command is output to increase the instantaneous rotation speed. If the instantaneous speed is higher than the target standby speed and the difference between the two exceeds the control dead zone, a zero power command is output to cause the flywheel energy storage unit to glide and decelerate naturally. If the difference between the current instantaneous speed and the target standby speed is within the control dead zone, a power command is output to offset standby losses in order to maintain the instantaneous speed constant.

8. The control method based on the coordinated optimization of flywheel energy storage and supercapacitor as described in claim 6, characterized in that, When the feedforward combat readiness standby mode is activated, the specific execution process of step three also includes: Monitor the instantaneous terminal voltage of the supercapacitor unit; If the instantaneous terminal voltage is lower than the preset maximum operating voltage, a charging power command is generated for the supercapacitor unit to raise the instantaneous terminal voltage to the maximum operating voltage or a preset voltage value close to the maximum operating voltage, so as to reduce the initial uniform marginal cost of the supercapacitor unit.

9. The control method based on the coordinated optimization of flywheel energy storage and supercapacitor as described in claim 1, characterized in that, The short-term load prediction signal comes from data calculated by the upper-level energy management system based on the scheduling strategy, or from data predicted by an independent load prediction unit based on historical operating data and a time series analysis model.

10. The control method based on the coordinated optimization of flywheel energy storage and supercapacitor as described in claim 1, characterized in that, The specific process of converting the power commands of the flywheel energy storage unit and the supercapacitor unit into control signals executable by the underlying power conversion system in step four includes: The power command of the supercapacitor unit is divided by the instantaneous terminal voltage to obtain the current reference value, and the current reference value is sent to the power conversion system of the supercapacitor unit. The torque reference value is obtained by dividing the power command of the flywheel energy storage unit by the mechanical angular velocity corresponding to the instantaneous rotational speed, and the torque reference value is sent to the motor driver of the flywheel energy storage unit.

Images