Charging control method and system of energy storage battery

Abstract

The invention relates to the technical field of power electronics and energy storage system control, in particular to a charging control method and system for an energy storage battery, and the method comprises the following steps: S100, constructing a system dynamic prediction model; s200, system operation parameters are collected, and the internal state of the storage battery pack is estimated; s300, based on the system dynamic prediction model and the internal state of the storage battery pack, constructing and solving an optimization problem which takes the trigger angle of the thyristor rectifier as an optimization variable and comprises the state constraint of the storage battery pack in a prediction time domain, and generating an optimal trigger angle control scheme; the state constraint comprises a charging current upper limit, a terminal voltage upper limit and a temperature rise rate upper limit; and S400, outputting the optimal trigger angle corresponding to the current moment in the optimal trigger angle control scheme to a driving circuit of the thyristor rectifier, and returning to execute the S200. By adopting the scheme, power grid fluctuation can be actively adapted, multi-constraint safety control is realized, and undisturbed smooth transition from a constant current mode to a constant voltage mode is completed.

Description

Technical Field

[0001] This invention relates to the field of power electronics and energy storage system control technology, and in particular to a charging control method and system for energy storage batteries. Background Technology

[0002] In modern industrial automation and power systems, variable frequency drives (VFDs) are widely used as core power equipment. To ensure production continuity and prevent equipment shutdown due to voltage dips (voltage fluctuations) or momentary power outages, a DC stabilization system is typically configured on the DC bus side of the VFD. This type of system often uses a highly reliable multi-pulse (e.g., 12-pulse) thyristor rectifier as the power conversion unit. Through a rectifier transformer, it converts AC grid energy into DC power, both to charge backup battery banks and to supplement the DC bus voltage when the grid is normal.

[0003] In such systems, high-rate battery packs are key energy storage components. To ensure they can provide sufficient instantaneous support energy during grid failures and quickly recharge after grid restoration, the charging control system must possess high dynamic response accuracy and stringent safety management capabilities. Currently, battery charging control in DC stabilization systems generally adopts a traditional dual-closed-loop architecture based on proportional-integral (PI) regulators, operating according to a preset mode of "constant current (CC) charging → reaching voltage threshold → switching to constant voltage (CV) charging".

[0004] However, this traditional PI control strategy has gradually revealed the following significant technical limitations in practical industrial applications:

[0005] First, it has poor adaptability to wide-range fluctuations in the power grid: the voltage fluctuations in the power grid at industrial sites are severe and frequent. The parameters of traditional PI controllers are usually set for rated operating conditions. When the grid voltage deviates significantly from the rated value, the fixed PI parameters are difficult to maintain stable tracking of the charging current or voltage, which can easily lead to increased current ripple, slower adjustment speed, or even oscillation, affecting charging efficiency and system stability.

[0006] Secondly, it cannot explicitly handle the multiple "hard constraints" of the system: high-rate batteries have extremely strict safety limits on parameters such as charging current, cutoff voltage, and temperature rise rate. Traditional PI control is essentially linear feedback, which can only limit the control quantity retrospectively through output limiting, and cannot directly and proactively incorporate the above physical constraints at the controller design level. This leads to current overshoot and voltage overshoot during system startup, sudden load changes, or mode switching, which not only threatens battery safety but also accelerates its capacity decay and shortens its service life.

[0007] Finally, the dynamic state and thermal management of the battery's internal state are neglected: existing strategies typically rely solely on feedback control based on battery terminal voltage and current, failing to consider internal polarization effects, the time-varying characteristics of state of charge (SOC), and the significant temperature rise under high-current charging. Especially for high-rate lead-acid batteries, continuous high-current charging generates substantial heat. If battery temperature or its rate of rise is not considered a core control constraint, it can easily lead to thermal runaway, accelerating plate sulfation and aging, posing a significant safety hazard to the system.

[0008] Therefore, for voltage management systems using thyristor rectifiers, there is an urgent need for a control method that can actively adapt to grid fluctuations, explicitly handle multiple safety constraints such as current, voltage, and temperature, and achieve smooth and optimized charging based on the internal state of the battery, in order to solve the fundamental shortcomings of traditional PI control in high-performance and high-safety scenarios. Summary of the Invention

[0009] This invention provides a charging control method and system for energy storage batteries, which can actively adapt to grid fluctuations, achieve multi-constraint safety control, and complete a smooth transition from constant current to constant voltage mode.

[0010] The basic solution provided by this invention is as follows:

[0011] A charging control method for an energy storage battery, applied to a voltage management system including a thyristor rectifier and a battery pack, includes the following steps:

[0012] S100, Construct a system dynamic prediction model; the system dynamic prediction model is used to predict the operating status of the thyristor rectifier and the battery pack in the prediction time domain;

[0013] S200: Collect system operating parameters and analyze the internal state of the battery pack based on a state estimation algorithm;

[0014] S300, based on the system dynamic prediction model and the internal state of the battery pack, construct and solve an optimization problem with the firing angle of the thyristor rectifier as the optimization variable and including the state constraints of the battery pack in the prediction time domain, and generate the optimal firing angle control scheme; the state constraints include the upper limit of charging current, the upper limit of terminal voltage, and the upper limit of temperature rise rate;

[0015] S400, the optimal firing angle corresponding to the current moment in the optimal firing angle control scheme is output to the drive circuit of the thyristor rectifier, and then the process returns to execute S200.

[0016] Furthermore, the predicted operating states in S100 include the output voltage of the thyristor rectifier, the charging current of the battery pack, and the terminal voltage of the battery pack.

[0017] Furthermore, the internal states analyzed in S200 include the state of charge and polarization voltage of the battery pack.

[0018] Furthermore, the system operating parameters include grid voltage, battery pack terminal voltage, battery pack charging current, and battery pack temperature.

[0019] Furthermore, the state estimation algorithm is the extended Kalman filter algorithm.

[0020] Furthermore, the battery pack is a high-rate energy storage battery pack.

[0021] The second basic solution provided by this invention:

[0022] A charging control system for an energy storage battery includes a thyristor rectifier, a battery pack, a drive module, a sensing module, and a control module.

[0023] The AC side of the thyristor rectifier is connected to the power grid, and the DC side is connected to the battery pack for charging the battery pack.

[0024] The drive module is connected to the thyristor rectifier and is used to adjust the firing angle of the thyristor rectifier;

[0025] The sensing module is used to collect system operating parameters;

[0026] The control module is connected to the drive module and the sensing module respectively, and is configured to execute the above-described charging control method for the energy storage battery.

[0027] Furthermore, the control module includes a state estimation unit and a model prediction control unit;

[0028] The state estimation unit is used to analyze the internal state of the battery pack based on the system operating parameters collected by the sensing module.

[0029] The model prediction control unit is used to construct and solve an optimization problem based on the system dynamic prediction model and the internal state of the battery pack, with the firing angle of the thyristor rectifier as the optimization variable and including the state constraints of the battery pack in the prediction time domain, and output the optimal firing angle to the drive module.

[0030] The principles and advantages of this invention are as follows:

[0031] First, by establishing a dynamic prediction model and employing a rolling optimization mechanism, this scheme can dynamically reconstruct and solve the optimization problem within each control cycle based on real-time collected system operating parameters. This allows it to proactively adapt to wide-range fluctuations in grid voltage, effectively suppress the impact of disturbances on the charging process, significantly improve the dynamic response speed and overall robustness of the voltage management system, and overcome the inherent defect of traditional fixed-parameter PI control in terms of deteriorating regulation quality under varying operating conditions.

[0032] Secondly, this scheme directly uses the upper limits of charging current, terminal voltage, and temperature rise rate as constraints in the optimization problem. This allows for proactive prediction and avoidance of any state that violates the safety boundaries when solving for the optimal firing angle control scheme. Therefore, it fundamentally eliminates the risks of current overshoot, voltage overshoot, and thermal runaway, providing the battery pack with safety protection throughout the entire charging process and extending battery life.

[0033] Furthermore, this solution uses a state estimation algorithm to estimate the internal state of the battery pack online and feeds it back to the system's dynamic prediction model, which improves the model's accuracy under varying operating conditions. This allows control decisions to closely align with the battery's actual physical characteristics and dynamic processes, thereby optimizing the charging trajectory and improving charging efficiency and control precision while ensuring safety.

[0034] Finally, based on the predictive time-domain optimization and state constraint handling capabilities of the model predictive control framework, this solution can automatically and smoothly adjust the charging current according to the predicted value of the battery pack terminal voltage, completing a natural transition from constant current to constant voltage charging mode without the need for external switching logic. The entire process is free of current or voltage surges, ensuring the stability of the charging process and equipment safety.

[0035] In summary, this solution achieves a comprehensive improvement in adaptability, safety, accuracy, and stability, and solves the fundamental shortcomings of traditional PI control in high-performance and high-safety scenarios. Attached Figure Description

[0036] Figure 1 This is a flowchart illustrating an embodiment of a charging control method for an energy storage battery according to the present invention.

[0037] Figure 2 This is a structural block diagram of the voltage management system in an embodiment of the charging control method for an energy storage battery according to the present invention.

[0038] Figure 3 The output voltage waveform during the system startup process using a traditional PI control strategy is shown in the experimental waveform diagram.

[0039] Figure 4 The output voltage waveform during the system startup process is shown in the figure.

[0040] Figure 5 This is a schematic diagram of the voltage and current response curves during the charging process of an energy storage battery charging control method according to the present invention, showing a seamless and smooth transition from constant current to constant voltage. Detailed Implementation

[0041] The following detailed description illustrates the specific implementation method:

[0042] Example 1:

[0043] A charging control method for an energy storage battery is disclosed, applied to a voltage regulation system comprising a thyristor rectifier and a battery pack. This embodiment uses a 20kW voltage regulation device as an example to implement the control method of the present invention. The battery pack is a high-rate energy storage battery pack; the thyristor rectifier is a multi-channel thyristor rectifier, specifically a 12-pulse thyristor rectifier. Figure 1 As shown, the method includes the following steps:

[0044] S100, Construct a system dynamic prediction model; the system dynamic prediction model is used to predict the operating state of the thyristor rectifier and battery pack in the prediction time domain. In this embodiment, the predicted operating state includes the output voltage of the thyristor rectifier, the charging current of the battery pack, and the terminal voltage of the battery pack. Specifically, the system dynamic prediction model is composed of a coupled second-order RC equivalent circuit model of the battery, a battery state of charge (SOC) change model, a battery temperature dynamic model, and a 12-pulse thyristor rectifier model. The structural block diagram of the voltage management system is as follows. Figure 2 As shown, it includes a three-phase power grid system, a multi-channel thyristor rectifier, a high-rate battery pack, and a phase-shifting transformer. The AC side of the multi-channel thyristor rectifier is connected to the three-phase power grid system through the phase-shifting transformer, while the DC side charges the high-rate battery pack.

[0045] The second-order RC equivalent circuit model of the battery is used to describe the dynamic characteristics of the internal polarization and ohmic resistance of the battery pack. The mathematical expression is as follows:

[0046]

[0047] In the formula, Let V be the terminal voltage of the battery pack at time k; The open-circuit voltage of the battery pack at time k is in V and can be obtained from the OCV-SOC characteristic curve provided by the battery manufacturer. The state of charge of the battery pack at time k, in % %. The internal resistance of the battery pack is expressed in ohms (Ω). The charging current of the battery pack at time k is expressed in A. and These are the polarization voltages of the two RC polarization branches, in V. , , , These represent the resistance (Ω), capacitance (F), and resistance (Ω), and capacitance (F) of the first RC polarization branch of the battery pack, respectively. The system sampling period, in seconds; , , , These represent the polarization voltages of the first RC polarization branch of the battery pack at time k+1, the first RC polarization branch of the battery pack at time k, the second RC polarization branch of the battery pack at time k+1, and the second RC polarization branch of the battery pack at time k, respectively, in V.

[0048] The battery pack state of charge (SOC) change model uses the ampere-hour integral method combined with open-circuit voltage correction, and its mathematical expression is as follows:

[0049]

[0050] In the formula, The state of charge of the battery pack at time k+1, in % %. For Coulomb efficiency; This refers to the rated capacity of the battery pack, in Ah.

[0051] The dynamic temperature model of the battery pack is established based on the thermal balance equation, and its mathematical expression is as follows:

[0052]

[0053] In the formula, This represents the temperature of the battery pack at time k+1, in °C. This represents the temperature of the battery pack at time k, in °C. This indicates the mass of the battery pack, in kg. This indicates the specific heat capacity of the battery pack, expressed in J / (kg·℃). This indicates the total internal resistance of the battery pack, in Ω. This indicates the heat dissipation coefficient of the battery pack, with units of W / (m²·℃). This indicates the heat dissipation area of ​​the battery pack, in m². Indicates ambient temperature, unit: ℃.

[0054] The 12-pulse thyristor rectifier model characterizes the nonlinear relationship between the firing angle and the DC output voltage, and its mathematical expression is as follows:

[0055]

[0056] In the formula, The DC output voltage of the rectifier, in volts (V). The effective value of the phase voltage of the power grid, in V; The firing angle of the thyristor is expressed in rad.

[0057] S200: Collect system operating parameters and analyze the internal state of the battery pack based on a state estimation algorithm. The system operating parameters include grid voltage, battery pack terminal voltage, battery pack charging current, and battery pack temperature. The state estimation algorithm is an extended Kalman filter algorithm, and the analyzed internal state includes the battery pack's state of charge and polarization voltage. In this embodiment, the extended Kalman filter algorithm specifically uses the battery pack's charging current as input and the battery pack terminal voltage as the observed value to analyze the battery pack's state of charge and polarization voltage. and It performs joint online estimation and updating of the model's internal resistance and ohmic resistance, providing accurate initial values ​​of the model state in real time.

[0058] S300, based on the system dynamic prediction model and the internal state of the battery pack, constructs and solves an optimization problem with the firing angle of the thyristor rectifier as the optimization variable and including the state constraints of the battery pack in the prediction time domain, generating an optimal firing angle control scheme; the state constraints include an upper limit for charging current, an upper limit for terminal voltage, and an upper limit for temperature rise rate. S300 includes:

[0059] S301, Constructing a multi-objective performance function:

[0060]

[0061] In the formula, This represents the performance function value at time k, used to evaluate the control effect; the smaller the value, the better the control effect. This indicates the prediction time domain length, i.e., how many sampling periods are predicted in the future; This indicates the length of the control time domain, i.e., how many future control actions are optimized; This represents the predicted output of the system at time k+i at time k. It is the charging current during the constant current charging stage and the terminal voltage during the constant voltage charging stage. This represents the output reference trajectory at time k+i, which is the set charging current value during the constant current charging stage and the set terminal voltage value during the constant voltage charging stage. This represents the control increment sequence at time k+j, i.e., the change in the firing angle; This indicates the predicted temperature rise, which is related to the real-time temperature of the battery pack. This represents the temperature rise at the end of the predicted time domain, i.e. The difference between the temperature at a given moment and the current temperature; Q represents the weight matrix of the output tracking error, used to adjust the importance of output tracking; R represents the weight matrix of the control increment change, used to adjust the control smoothness; This represents the weighting coefficient of the temperature rise suppression term, used to adjust the importance of temperature control.

[0062] S302, Set system state constraints:

[0063]

[0064] In the formula, and These represent the minimum and maximum allowable charging currents of the battery pack, respectively, in A. and These represent the minimum and maximum cutoff voltages of the battery pack, respectively, in V; and These represent the lower and upper limits of the safe state of charge of the battery pack, respectively, in % %. and These represent the minimum and maximum values ​​of the thyristor firing angle, respectively, in rad. This indicates the maximum permissible rate of temperature rise of the battery pack, in °C / s.

[0065] S303, an online solution for constrained quadratic programming problems. Within each control cycle, the multi-objective optimization problem is transformed into a standard quadratic programming form:

[0066]

[0067] In the formula, Indicates the future Control increment sequence at time 1 H represents the Hessian matrix, which is derived from the system state-space matrix and weight matrices Q and R, and describes the coefficients of the quadratic terms. This represents a vector of linear terms, describing the coefficients of the linear terms.

[0068] The effective set method is used to solve this quadratic programming problem online, yielding the optimal control increment sequence ∆U*. An automatic smooth transition mechanism from constant current to constant voltage charging mode is implemented: In the initial stage of charging, when the predicted battery pack terminal voltage V... bat Below the maximum cutoff voltage V max At this time, the optimizer primarily tracks the constant current reference value; as charging progresses, when the predicted terminal voltage V... bat Approaching or reaching Vmax When constraining the boundary conditions, the quadratic programming solver automatically and smoothly reduces the charging current to satisfy the hard constraint, thus achieving a seamless transition from constant current charging mode to constant voltage charging mode without the need for external switching logic. The effect of this mechanism is as follows: Figure 5 As shown, Figure 5 The voltage and current response curves during charging using the method of this invention are shown. The constant current charging current is set to 5A, and the maximum charging voltage is 535V. It can be seen that when the charging voltage is below 535V, the system operates in constant current charging mode; when the voltage reaches the 535V constraint boundary, the charging current smoothly decreases under the control of the optimizer, and the system automatically and seamlessly transitions to constant voltage charging mode, without any voltage or current surges throughout the process.

[0069] S400: Output the optimal firing angle corresponding to the current moment in the optimal firing angle control scheme to the drive circuit of the thyristor rectifier, and return to execute S200. Specifically, the first element of the optimal control increment sequence ∆U* obtained in step S300, i.e., the optimal firing angle increment at the current moment, is... Acting on the system; calculating the thyristor firing angle at the current moment. The output is then sent to the drive circuit. In the next sampling cycle, the system state is updated, and S200 to S400 are repeated to achieve rolling optimization closed-loop control.

[0070] To verify the superiority of the method of this invention, comparative experiments were conducted. (See attached document.) Figure 3 The experimental waveform of the output voltage during system startup is shown when using a traditional PI control strategy. It is evident that due to the fixed PI parameters and lack of constraint prediction, significant transient oscillations occur in the output voltage during the initial startup phase, with voltage spikes reaching as high as 166V, posing a safety hazard. (Appendix) Figure 4 The startup output voltage waveform is shown when the model predictive control strategy described in this invention is employed. As can be seen from the comparison, thanks to the explicit handling of output constraints by the MPC, the output voltage rises smoothly without overshoot or overcharge, effectively eliminating startup shock and verifying the significant advantages of this invention in dynamic response and safety.

[0071] Example 2:

[0072] A charging control system for an energy storage battery includes a thyristor rectifier, a battery pack, a drive module, a sensing module, and a control module.

[0073] The AC side of the thyristor rectifier is connected to the power grid, and the DC side is connected to the battery pack for charging the battery pack.

[0074] The drive module is connected to the thyristor rectifier and is used to adjust the firing angle of the thyristor rectifier; specifically, it adjusts the firing phase of each thyristor in the thyristor rectifier according to the received firing angle control signal.

[0075] The sensing module is used to collect system operating parameters, including: grid voltage, grid current, battery pack terminal voltage, battery pack charging current, and battery pack temperature.

[0076] The control module is connected to both the drive module and the sensing module, and is configured to execute the charging control method for the energy storage battery as described in Embodiment 1. The control module includes a state estimation unit and a model prediction control unit.

[0077] The state estimation unit is used to analyze the internal state of the battery pack based on the system operating parameters collected by the sensing module. Specifically, it receives the system operating parameters collected by the sensing module, runs the extended Kalman filter algorithm, and uses the charging current as input and the battery pack terminal voltage as observation to perform real-time joint estimation of the state of charge, polarization voltage, and ohmic internal resistance of the battery pack, providing accurate initial state values ​​for the prediction model.

[0078] The model prediction control unit is used to construct and solve an optimization problem based on the system dynamic prediction model and the internal state of the battery pack. This problem involves the firing angle of the thyristor rectifier as the optimization variable and includes state constraints of the battery pack within the prediction time domain. The unit then outputs the optimal firing angle to the drive module. Specifically, the model prediction control unit stores and executes the system dynamic prediction model as described in step S100 of Example 1. In each control cycle, based on the output of the state estimation unit and the currently collected system operating parameters, this unit constructs a multi-objective performance function as described in step S300 of Example 1 and applies multiple state constraints to form a constrained quadratic programming problem. By solving this quadratic programming problem, an optimal firing angle sequence is generated, and the optimal firing angle at the current moment is output to the drive module, achieving safe, smooth, and efficient charging closed-loop control.

[0079] By adopting the above scheme, the charging control system can achieve precise, safe, and efficient charging management of high-rate energy storage battery packs under grid voltage fluctuation conditions, effectively extending the service life of the battery packs.

[0080] The above are merely embodiments of the present invention. Commonly known structures and characteristics are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A method for controlling the charging of an energy storage battery, applied to a voltage management system comprising a thyristor rectifier and a battery bank, characterized in that: The method comprises the following steps: S100, constructing a system dynamic prediction model; the system dynamic prediction model is used to predict the running state of the thyristor rectifier and the battery pack in a prediction time domain; S200, collecting system running parameters and analyzing the internal state of the battery pack based on a state estimation algorithm; S300, based on the system dynamic prediction model and the internal state of the battery pack, constructing and solving an optimization problem with the trigger angle of the thyristor rectifier as an optimization variable and containing state constraints of the battery pack in the prediction time domain, to generate an optimal trigger angle control scheme; the state constraints include an upper limit of the charging current, an upper limit of the terminal voltage and an upper limit of the temperature rise rate; S400, outputting the optimal trigger angle corresponding to the current time in the optimal trigger angle control scheme to the driving circuit of the thyristor rectifier, and returning to execute S200.

2. The charge control method of an energy storage battery according to claim 1, characterized by: The running state predicted in S100 includes the output voltage of the thyristor rectifier, the charging current of the battery pack and the terminal voltage of the battery pack.

3. The charge control method of an energy storage battery according to claim 1, characterized by: The internal state analyzed in S200 includes the state of charge and the polarization voltage of the battery pack.

4. The charge control method of an energy storage battery according to claim 1, characterized by: The system running parameters include the grid voltage, the terminal voltage of the battery pack, the charging current of the battery pack and the temperature of the battery pack.

5. The charge control method of an energy storage battery according to claim 1, characterized by: The state estimation algorithm is an extended Kalman filter algorithm.

6. The charge control method of an energy storage battery according to claim 1, characterized by: The battery pack is a high-rate energy storage battery pack.

7. A charge control system for an energy storage battery, characterized by: The system comprises a thyristor rectifier, a battery pack, a driving module, a sensing module and a control module; The AC side of the thyristor rectifier is connected to the grid, and the DC side is connected to the battery pack, for charging the battery pack; The driving module is connected to the thyristor rectifier, for adjusting the trigger angle of the thyristor rectifier; The sensing module is used to collect system running parameters; The control module is connected with the driving module and the sensing module respectively, and is configured to execute the charging control method of the energy storage battery according to any one of claims 1 to 6.

8. A charge control system for an energy storage battery as claimed in claim 7, characterised in that: The control module comprises a state estimation unit and a model prediction control unit; The state estimation unit is used to analyze the internal state of the battery pack based on the system running parameters collected by the sensing module; The model prediction control unit is used to construct and solve an optimization problem with the trigger angle of the thyristor rectifier as an optimization variable and containing state constraints of the battery pack in the prediction time domain based on the system dynamic prediction model and the internal state of the battery pack, and output the optimal trigger angle to the driving module.

Images