A power distribution network transformer and cascaded voltage reduction device combined voltage regulation method and system

Abstract

The application discloses a power distribution network transformer and cascaded voltage reduction device combined voltage regulation method and system, relates to the field of intelligent power distribution network voltage control, and comprises the following steps: in response to the load side output voltage deviating from the target reference value, a global comprehensive loss model containing transformer and cascaded voltage reduction device loss is constructed; based on the model, candidate gear expected total loss is evaluated, a cost function is constructed by combining wear and tear penalty weight to determine the target gear; when the anti-shake interception condition is met, a switching instruction is issued; during the gear shifting process, the actual input voltage is monitored, when a step mutation is detected, the closed-loop feedback is suspended and the transient feedforward direct control mode is switched in to update the driving signal; after the voltage is stable, the closed-loop control is restored. The application realizes the deep cooperation of macroscopic coarse adjustment and microscopic fine adjustment and the seamless connection of gear shifting transient state, and improves the system efficiency and power supply quality.

Description

Technical Field

[0001] This invention belongs to the field of intelligent distribution network voltage control, specifically relating to a method and system for combined voltage regulation of distribution network transformers and cascaded step-down devices. Background Technology

[0002] Currently, in the field of voltage regulation in distribution networks, on-load tap-changing transformers are commonly used as the basic voltage regulating equipment. Voltage regulation is achieved by changing the turns ratio through mechanical switching of tap positions. However, this method of voltage regulation relying on mechanical contact action suffers from drawbacks such as slow response speed, discrete step-like regulation, and frequent operation that easily damages the mechanical contacts. This results in an inability to achieve fast, smooth, and high-precision voltage control when facing high-frequency voltage fluctuations caused by distributed power source access or transient voltage drops caused by impulsive loads. Although existing technologies have introduced power electronic converters to assist in voltage regulation, the simple combination scheme lacks a global energy efficiency coordination mechanism, and the voltage step generated at the moment of transformer mechanical tap switching can cause severe transient impacts on downstream outputs, affecting power quality and equipment lifespan. Therefore, there is an urgent need for a combined voltage regulation scheme that can achieve deep decoupling between slow coarse adjustment and high-speed fine adjustment, while balancing optimal global efficiency and seamless transient response. Summary of the Invention

[0003] To address the shortcomings of existing technologies, the present invention aims to provide a method and system for joint voltage regulation of distribution network transformers and cascaded step-down devices, which realizes deep synergy between macroscopic discrete voltage regulation and microscopic continuous voltage regulation, optimization of global operating efficiency, and active elimination of shift transient impacts.

[0004] This invention provides a method for combined voltage regulation of a distribution network transformer and a cascaded step-down device, comprising:

[0005] In response to the load-side output voltage deviating from the target output voltage reference value, a global comprehensive loss model is constructed, which includes the losses of the on-load tap-changing transformer and the losses of the cascaded step-down device.

[0006] Based on the global comprehensive loss model, the expected total system loss corresponding to multiple candidate taps of the on-load tap-changing transformer is evaluated, and a cost function is constructed by combining the equipment wear penalty weight to determine the target tap.

[0007] Under the condition that the preset anti-shake interception conditions are met, a control command to switch to the target tap is sent to the on-load tap-changing transformer.

[0008] During the tap change of the on-load tap-changing transformer, monitor the actual input voltage of the cascaded step-down device;

[0009] In response to the detection of a step change in the actual input voltage, the voltage closed-loop feedback control is suspended, and the transient feedforward direct control mode is switched to. The feedforward duty cycle is calculated based on the actual input voltage after the step change and the target output voltage reference value, and the drive signal of the cascaded buck converter is updated.

[0010] In response to the stable operation of the actual input voltage, voltage closed-loop feedback control is restored.

[0011] A global comprehensive loss model is constructed, incorporating losses from on-load tap-changing transformers and cascaded step-down devices. This includes: constructing a comprehensive loss evaluation function covering losses from on-load tap-changing transformers, conduction losses from cascaded step-down devices, switching losses, and filter inductor losses; evaluating the expected total system loss corresponding to multiple candidate tap positions of the on-load tap-changing transformer, including: for each candidate tap position, calculating the predicted input voltage of the cascaded step-down device when switching to that candidate tap position based on the grid-side background voltage and load current; and substituting the predicted input voltage into the comprehensive loss evaluation function to calculate the expected total system loss corresponding to maintaining rated output when switching to that candidate tap position.

[0012] The formula for predicting the input voltage is: In the formula, To switch to the 1st Predicted input voltage at gear position The background voltage is collected in real time from the power grid side. For on-load tap-changing transformers The rated no-load output voltage of the gear position. For load current, For the first Equivalent short-circuit impedance at the specified gear position.

[0013] A cost function is constructed by combining equipment wear penalty weights. This includes: combining the expected total system loss with a penalty term representing the equipment wear cost caused by mechanical shifting to construct a cost function for evaluating the comprehensive economic efficiency of candidate shift points, in order to avoid frequent switching between near-optimal solutions due to grid voltage fluctuations; under the condition of meeting preset anti-jitter interception conditions, a control command to switch to the target shift point is issued to the on-load tap-changing transformer, including: performing the first level of safety interception judgment: verifying the difference between the target shift point and the current shift point and the current grid voltage change trend. If the difference between the target shift point and the current shift point is determined to be one shift point and the current grid voltage change trend is reversing, the issuance of the control command is temporarily suspended; performing the second level of safety interception judgment: reading the interval time since the last shift point action. If the interval time is less than the preset mechanical protection time threshold, the control command is blocked; if it is greater than or equal to the mechanical protection time threshold, the control command is officially issued.

[0014] The formula for calculating the cost function is: In the formula, For the first The comprehensive cost function value of each candidate gear position. To switch to the 1st The predicted total system loss at gear position. This is the weighting coefficient for equipment wear penalty. This refers to the tap number of the on-load tap-changing transformer currently under consideration. This refers to the current actual tap position of the on-load tap-changing transformer; the control command is issued to the target tap position corresponding to the tap position number that minimizes the comprehensive cost function value.

[0015] In response to the detection of a step change in the actual input voltage, the voltage closed-loop feedback control is suspended, and the transient feedforward direct control mode is switched on. The feedforward duty cycle is calculated based on the actual input voltage after the step change and the target output voltage reference value, and the drive signal of the cascaded buck converter is updated. This includes: instantaneously suspending the conventional closed-loop feedback network with integral hysteresis within the microsecond-level control cycle in which the step change in the actual input voltage is detected; directly acquiring the transient actual input voltage and recalculating the feedforward duty cycle using the transient feedforward formula; converting the feedforward duty cycle into the corresponding timer count value and writing it into the comparison register of the pulse width modulation module so as to update the drive signal within one pulse width modulation switching cycle.

[0016] The transient feedforward formula is expressed using the following equation: ;in, V is the feedforward duty cycle; ref V is the target output voltage reference value. in_step I represents the actual input voltage after a step jump. load R is the load current. eq This is the equivalent parasitic resistance of the main circuit of the cascaded step-down device.

[0017] In response to the stable operation of the actual input voltage, the voltage closed-loop feedback control is restored, including: after determining that the actual input voltage step has ended and is operating stably at the voltage corresponding to the target level, the transient feedforward direct control mode is smoothly exited, and the voltage dual closed-loop feedback control network containing the voltage outer loop and the current inner loop is restored; wherein, the voltage outer loop is used to eliminate the steady-state error of the system, and the current inner loop is used to provide system damping and current limiting protection to ensure that the cascaded step-down device continuously outputs the rated voltage to the load that meets the target output voltage reference value.

[0018] In response to the load-side output voltage deviating from the target output voltage reference value, a global comprehensive loss model is constructed, including the losses of the on-load tap-changing transformer and the cascaded step-down device. This model includes: setting the target output voltage reference value and the voltage dead zone threshold for allowable fluctuations; when the absolute value of the difference between the load-side output voltage and the target output voltage reference value is detected to be greater than the voltage dead zone threshold for N consecutive power frequency cycles, the construction of the global comprehensive loss model and the determination of the target voltage level are triggered, where N is a positive integer; when the absolute value of the difference is determined to be less than or equal to the voltage dead zone threshold, it is determined that the current state is in the optimal steady state and the current operating state is maintained.

[0019] This invention also provides a combined voltage regulation system for a distribution network transformer and a cascaded step-down device, comprising: a main power unit, including an on-load tap-changing transformer, a cascaded step-down device, and an output filter network connected in series; a status detection network, including a grid voltage sensor for monitoring the background voltage of the distribution network, an intermediate voltage sensor for acquiring the actual input voltage of the cascaded step-down device, a load-side sensor for acquiring the load-side output voltage and load current, and an inductor current sensor for acquiring the inductor current inside the cascaded step-down device; a main control unit, connected to the main power unit and the status detection network, configured to execute the combined voltage regulation method as described above; and a bypass protection unit, connected to the main control unit, configured to block the drive pulse and close the bypass switch when a fault is detected in the internal components of the cascaded step-down device or an extreme drop in grid voltage, so that the on-load tap-changing transformer directly supplies power to the load, bypassing the cascaded step-down device.

[0020] The aforementioned system provides a reliable hardware execution platform for the joint voltage regulation method through multi-dimensional state perception and collaboration with the main control unit, and ensures power supply continuity under extreme operating conditions through a bypass protection mechanism.

[0021] In addition, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the combined voltage regulation method as described above.

[0022] This invention discloses a method and system for joint voltage regulation of a distribution network transformer and a cascaded step-down device. By constructing a global comprehensive loss model covering various losses of the transformer and the cascaded step-down device, and combining equipment wear penalty weights to construct a cost function for optimal voltage regulation, this method fundamentally breaks down the traditional isolated control of equipment, achieving deep decoupling and energy efficiency synergy between slow coarse adjustment and high-speed fine adjustment. Compared to existing technologies, this scheme can dynamically limit the input voltage of the cascaded step-down device to the minimum reasonable range that meets voltage regulation requirements, significantly reducing voltage stress on semiconductor devices and high-frequency inductor heating, and extending equipment lifespan exponentially. Simultaneously, it innovatively introduces a shift transient feedforward direct control mechanism. Within a microsecond-level window when a step change in input voltage caused by transformer shifting is detected, the conventional closed-loop network with integral hysteresis is instantaneously suspended and the underlying drive signal is directly refreshed. This eliminates voltage overshoot and drop caused by mechanical delay from the control mechanism perspective, perfectly masking the coarseness of discrete transformer voltage regulation, achieving absolute smoothness and seamless connection throughout the voltage regulation process, and effectively solving the problem of bidirectional voltage fluctuation and limit exceeding at the end of the distribution network. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the topology of the distribution network transformer and cascaded step-down device combined voltage regulation system according to an embodiment of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0026] This embodiment provides a method and system for joint voltage regulation of a distribution network transformer and a cascaded step-down device. This method establishes a hierarchical collaborative control logic that prioritizes micro-level voltage regulation by power electronics and provides macro-level voltage adjustment by the transformer, aiming to solve the problems of slow response and low accuracy in traditional voltage regulation methods, as well as the lack of energy efficiency synergy and transient suppression capabilities in hybrid voltage regulation systems. Specifically, the joint voltage regulation method includes the following steps:

[0027] In step S100, in response to the load-side output voltage deviating from the target output voltage reference value, a global comprehensive loss model is constructed, which includes the losses of the on-load tap-changing transformer and the losses of the cascaded step-down device.

[0028] Specifically, the purpose of constructing a global integrated loss model is to break down the isolated control islands of on-load tap-changing transformers and cascaded step-down devices in traditional control systems, and to achieve a comprehensive consideration of the efficiency of both levels of equipment. In a distribution network, on-load tap-changing transformers are responsible for macroscopic voltage level transformation, while cascaded step-down devices are responsible for microscopic precise adjustment; their loss characteristics are coupled. For example, when the tap position of the on-load tap-changing transformer is increased, although it may reduce its own copper losses, it will lead to an increase in the input voltage of the cascaded step-down device, thereby increasing its switching losses and inductor ripple losses. Therefore, this embodiment, by constructing a global model covering the physical characteristics of both levels of equipment, can accurately assess the total system energy consumption under different operating conditions, providing a precise physical benchmark for subsequent efficiency optimization and avoiding the problem of global energy efficiency deterioration caused by local optimization.

[0029] Step S200: Based on the global integrated loss model, evaluate the expected total system loss corresponding to multiple candidate taps of the on-load tap-changing transformer, and construct a cost function in combination with equipment wear penalty weights to determine the target tap.

[0030] In this step, the system iterates through all feasible candidate tap positions of the on-load tap-changing transformer and uses the aforementioned global comprehensive loss model to calculate the expected total system loss required to maintain rated output at each tap position. It is important to note that this loss assessment is a theoretical calculation based on the predicted input voltage, which is derived from the current grid background voltage, load current, and transformer turns ratio, and is used to characterize the steady-state operating point after switching to that tap position. Simultaneously, to avoid frequent mechanical switching of the transformer due to a pursuit of optimal efficiency, this embodiment introduces a device wear penalty weight in the cost function. The technical motivation is to balance the system's operational economy with equipment lifespan, explicitly quantifying the implicit wear cost of mechanical switching. For example, when the expected loss difference between two adjacent tap positions is extremely small, the penalty weight will lower the cost of maintaining the current tap position, thereby suppressing unnecessary switching operations and ensuring that the determined target tap position is both an energy-efficient advantage and a safe point for the equipment.

[0031] In step S300, if the preset anti-shake interception conditions are met, a control command to switch to the target tap is sent to the on-load tap-changing transformer.

[0032] In this embodiment, the anti-jitter interception mechanism plays a crucial safety barrier role. Its functional definition is to prevent the system from oscillating and frequently switching voltages in the critical voltage fluctuation region. Since the distribution network voltage itself exhibits random fluctuations, issuing commands based solely on the instantaneous optimal result can easily lead to repeated voltage level changes by the transformer within a short period. This not only fails to improve power quality but also accelerates arcing at mechanical contacts and may even cause equipment failure. Therefore, this embodiment employs strict interception logic, allowing commands to be issued only when the target voltage level switching demand is continuous and meets mechanical safety constraints. This design effectively filters out false voltage regulation demands caused by grid noise, ensuring the long-term reliable operation of the on-load tap-changing transformer.

[0033] Step S400: During the tap change of the on-load tap-changing transformer, monitor the actual input voltage of the cascaded step-down device.

[0034] This step is the crucial link between macroscopic shifting and microscopic compensation. Unlike the predicted input voltage used for optimization calculation in step S200, this step monitors the real-time physical value acquired by the sensor. Because the mechanical action of the on-load tap-changing transformer has a physical delay of hundreds of milliseconds, and unpredictable voltage steps occur at the instant the contacts open and close, this transient process cannot be accurately described by a mathematical model alone. By monitoring the actual input voltage at high frequency, the system can sensitively capture the real voltage disturbances caused by mechanical shifting, providing an accurate data source for subsequent rapid compensation.

[0035] In step S500, in response to the detection of a step change in the actual input voltage, the voltage closed-loop feedback control is suspended, and the transient feedforward direct control mode is switched to. The feedforward duty cycle is calculated based on the actual input voltage after the step change and the target output voltage reference value, and the drive signal of the cascaded buck converter is updated.

[0036] This is the core action in this embodiment to address transient shocks during gear shifts. When a step change in the actual input voltage is detected, conventional voltage closed-loop feedback control inherently has phase lag due to its integral component. If operation continues, the integrator will accumulate the error from the sudden change, leading to severe overshoot or prolonged settling time. Therefore, this embodiment adopts a strategy of instantaneously suspending the closed loop to prevent integral saturation. It then switches to transient feedforward direct control mode, an open-loop control mechanism that directly calculates the feedforward duty cycle using the ratio of the actual input voltage after the step change to the target output voltage reference value, and immediately updates the drive signal of the cascaded buck converter. This approach bypasses the bandwidth limitation of the closed-loop regulator, enabling a response to voltage steps on a microsecond timescale, canceling out input disturbances before they reach the output, thus achieving seamless transition during gear shifts.

[0037] In step S600, in response to the stable operation of the actual input voltage, the voltage closed-loop feedback control is restored.

[0038] While transient feedforward direct control offers a fast response, it is an open-loop control system. Limited by factors such as device parameter drift and sampling accuracy, it cannot completely eliminate steady-state errors. Therefore, once the actual input voltage completes a step change and enters a new steady-state operating range, closed-loop voltage feedback control must be promptly restored. Restoring the closed loop is necessary to utilize the infinite gain characteristic of the feedback mechanism to eliminate residual steady-state deviations from the feedforward control, ensuring that the load-side output voltage is maintained accurately near the target reference value over a long period. Simultaneously, closed-loop control provides necessary system damping and disturbance rejection capabilities to cope with subsequent load changes or minor grid fluctuations. Through this smooth switching between transient feedforward and steady-state closed-loop control, this embodiment achieves high-quality voltage regulation throughout all time periods.

[0039] Through the coordinated operation of steps S100 to S600, this embodiment establishes a methodological framework for the upper-level control of the joint voltage regulation system. This framework does not rely on specific hardware topologies or mathematical formulas, but rather achieves efficient management of voltage fluctuations in the distribution network by defining the logical relationships and timing coordination of each functional module. It should be understood that the execution order of the above steps can be adaptively adjusted according to the actual controller architecture in specific implementations. For example, the monitoring action in step S400 can be a background task throughout the entire voltage regulation process, not limited to the shifting period. Furthermore, the specific composition of the global integrated loss model, the mathematical expression of the cost function, the detailed criteria for anti-jitter interception, and the calculation details of transient feedforward will be further explained in subsequent embodiments, but these specific implementation methods should not be considered as limitations on the upper-level logical framework of this embodiment.

[0040] Furthermore, the details of constructing the global integrated loss model, the calculation method for predicting the input voltage, and the trigger boundary for optimal startup are explained in detail. A global integrated loss model is constructed, incorporating losses from on-load tap-changing transformers and cascaded step-down devices. This includes constructing a comprehensive loss evaluation function covering on-load tap-changing transformer losses, cascaded step-down device conduction losses, switching losses, and filter inductor losses. Specifically, this comprehensive loss evaluation function incorporates the energy dissipation mechanisms of the two core devices in the system into a unified quantification system. The on-load tap-changing transformer losses mainly consist of iron losses and copper losses. Iron losses depend on the grid voltage amplitude and change relatively little during tap adjustment, while copper losses are proportional to the square of the load current and the winding resistance at the current tap. The conduction losses of the cascaded step-down devices depend on the on-resistance of the power switch, the duty cycle, and the effective value of the load current, reflecting the Ohmic thermal effect of the device in the on-state. The switching losses of the cascaded step-down devices are directly related to the input voltage, load current, switching frequency, and the overlap time of the devices, and are the most significantly affected component by voltage stress in high-frequency power electronic converters. The filter inductor loss consists of DC copper loss, AC skin effect copper loss, and high-frequency hysteresis loss of the magnetic core, with the AC component being proportional to the square of the inductor current ripple. Through refined modeling of these four losses, the system can accurately capture the dynamic changes of each loss component at different speeds. In particular, it reveals the physical mechanism by which reducing the input voltage of the cascaded buck converter can simultaneously and significantly suppress switching losses and inductor ripple losses, thus providing a solid physical benchmark for optimizing global efficiency.

[0041] The evaluation assesses the expected total system loss corresponding to multiple candidate tap positions of the on-load tap-changing transformer. This includes, for each candidate tap position, calculating the predicted input voltage of the cascaded step-down device when switching to that tap position based on the grid-side background voltage and load current. Substituting the predicted input voltage into a comprehensive loss assessment function, the expected total system loss for maintaining rated output when switching to that candidate tap position is calculated. In this process, the predicted input voltage is a key intermediate variable connecting macroscopic tap position decision-making and microscopic loss assessment. It is not a sensor-measured value, but a theoretical steady-state value derived from circuit topology parameters and real-time operating conditions. By substituting the predicted input voltage corresponding to each candidate tap position into the aforementioned comprehensive loss assessment function, the system can predict the total energy consumption level of the entire combined tap-changing system while maintaining the target output voltage after switching to that tap. This pre-evaluation mechanism based on a prediction model allows tap position optimization to be completed without actual mechanical action, avoiding equipment wear and voltage fluctuation risks associated with trial-and-error adjustments.

[0042] Regarding the specific calculation of the predicted input voltage, this embodiment uses the following formula: In the formula, To switch to the 1st Predicted input voltage at gear position The background voltage is collected in real time from the power grid side. For on-load tap-changing transformers The rated no-load output voltage of the gear position. For load current, For the first The equivalent short-circuit impedance at the specified tap position. This formula essentially describes the external characteristic equation of the transformer secondary port, that is, subtracting the voltage drop caused by the load current flowing through the transformer's internal impedance from the ideal no-load voltage. The equivalent short-circuit impedance is... Accurate acquisition of this information is crucial for ensuring prediction accuracy. In practical implementation, The short-circuit impedance data can be obtained by consulting the transformer's factory test report, or an online parameter identification algorithm can be used to perform least-squares fitting and real-time updates using voltage and current samples from historical operating data to adapt to parameter drift caused by transformer aging or temperature changes. It should be understood that although this embodiment provides a specific linear voltage drop model, in other embodiments, if saturation characteristics or harmonic effects are considered, more complex nonlinear models or lookup table methods can be used to calculate and predict the input voltage, as long as they can characterize the mapping relationship between tap position, load, and input voltage.

[0043] Furthermore, to avoid unnecessary frequent optimization calculations due to transient disturbances or normal ripple, this embodiment also sets strict trigger boundaries. In response to a deviation of the load-side output voltage from the target output voltage reference value, a global comprehensive loss model is constructed, incorporating losses from the on-load tap-changing transformer and the cascaded step-down device. This includes setting the target output voltage reference value and a voltage dead zone threshold for allowable fluctuations. When the absolute value of the difference between the load-side output voltage and the target output voltage reference value is detected to be greater than the voltage dead zone threshold for N consecutive power frequency cycles, the construction of the global comprehensive loss model and the determination of the target voltage level are triggered, where N is a positive integer. When the absolute value of the difference is determined to be less than or equal to the voltage dead zone threshold, it is determined that the current operating state is optimal and maintained. Specifically, the setting of the voltage dead zone threshold e should comprehensively consider the national standard voltage deviation limit and the sensor's measurement noise level. For example, in a preferred embodiment, e can be set between 2V and 3V, ensuring both the user's perception of voltage quality and effectively filtering out millivolt-level sampling noise. The setting of the number of consecutive cycles N is used to distinguish between transient disturbances and persistent voltage deviations. For example, N can be set to 3 to 5 power frequency cycles, which means that only when the voltage over-limit state lasts for more than 60ms to 100ms is it considered a real demand to initiate macroscopic adjustment. This dual filtering mechanism blocks false triggering caused by short-term impacts such as lightning surges and large motor starts from the source, ensuring that the global comprehensive loss model is only activated under steady-state conditions that truly have voltage regulation benefits, significantly reducing the ineffective computing power consumption of the main control unit and the probability of mechanical wear of the transformer.

[0044] Furthermore, the decision-making logic and safe execution mechanism for gear selection are explained in detail. A cost function is constructed by combining equipment wear penalty weights. This includes combining the expected total system loss with a penalty term representing the equipment wear cost caused by mechanical gear shifting, to construct a cost function for evaluating the overall economic efficiency of candidate gears, thus avoiding frequent switching between near-optimal solutions due to grid voltage fluctuations. Specifically, simply achieving optimal efficiency is not equivalent to achieving global optimality in system operation, because the mechanical contact life of on-load tap-changing transformers is a finite and scarce resource. If only minimizing losses is the objective, when the grid voltage fluctuates slightly near the critical point between two gears, the algorithm may calculate that the loss difference between adjacent gears is minimal or even alternately dominant, leading to repeated gear shifting in a short period. This oscillating adjustment not only fails to bring substantial energy efficiency benefits but also drastically accelerates arc burn-out of mechanical contacts and spring fatigue. Therefore, this embodiment introduces a penalty term to transform abstract mechanical wear into a quantifiable cost that can be directly added to electrical energy loss, enabling the cost function to simultaneously reflect operational economy and equipment safety.

[0045] As a specific implementation method, the formula for calculating the cost function is: In the formula, For the first The comprehensive cost function value of each candidate gear position. To switch to the 1st The predicted total system loss at the gear position is calculated by the comprehensive loss assessment function in Example 2. This is the weighting coefficient for equipment wear penalty. This refers to the tap number of the on-load tap-changing transformer currently under consideration. This refers to the current actual tap position of the on-load tap-changing transformer. The control command is issued to the target tap position corresponding to the tap number that minimizes the comprehensive cost function value. In this formula, the penalty weighting coefficient... The dimension of is power (e.g., watt W), and its physical meaning is "the instantaneous power loss cost equivalent to a single gear shifting action". Regarding The value selection strategy should be determined comprehensively based on the transformer's rated capacity, mechanical structure lifespan, and electricity cost. For example, in a preferred embodiment, for a 630kVA distribution network transformer, It can be set between 50W / level and 100W / level. This means that the cost function only applies when the expected total system losses saved by switching to the new level exceed 50W to 100W. Only when the cost of maintaining the current gear is less than the cost of keeping it in gear will a shift decision be triggered. This design is equivalent to artificially setting up a "holding zone" centered on the current gear in the optimization space, effectively filtering out ineffective adjustment needs that, while theoretically beneficial, are insufficient to cover wear and tear costs. It should be understood that... It can be a preset fixed constant, or a variable that is dynamically adjusted based on the number of times the transformer has operated or its health status, such as increasing as the transformer's service life increases. This value is used to further extend the remaining lifespan of older equipment.

[0046] After determining the target speed, the system does not immediately issue a command, but must first pass a strict safety check. If the preset anti-jitter interception conditions are met, a control command to switch to the target speed is issued to the on-load tap-changing transformer, specifically involving the execution of two safety interception checks.

[0047] Step S301: Perform the first level of safety interception judgment: verify the difference between the target level and the current level, as well as the current grid voltage change trend. If the difference between the target level and the current level is determined to be one level and the current grid voltage change trend is reversing, then the issuance of control commands is temporarily suspended. This level of interception mainly targets short-term spikes or drops in grid voltage. For example, when the output of distributed photovoltaic power generation suddenly decreases due to cloud cover, causing the voltage in the distribution area to drop instantaneously and reach the low-level switching threshold, if the rate of change of grid voltage is detected at this time... The voltage has turned from negative to positive (i.e., the voltage is recovering), indicating that the voltage deviation is a transient disturbance rather than a persistent offset. At this point, even if the cost function calculates a switch to a lower voltage level... At the very least, the system will proactively pause issuing commands, allowing the cascaded step-down devices with faster response times to continue handling the temporary differential pressure regulation. This trend-based soft-interception mechanism prevents the transformer from overreacting to transient disturbances at the source.

[0048] Step S302: Perform the second layer of safety interception judgment: Read the interval time since the last tap change action. If the interval time is less than the preset mechanical protection time threshold, the control command is blocked; if it is greater than or equal to the mechanical protection time threshold, the control command is officially issued. This layer of interception is a hard protection constraint on the mechanical structure of the on-load tap-changing transformer. Because the mechanical switch requires sufficient time for contact reset, insulating oil arc extinguishing, and transmission mechanism return to position after completing a tap change, forcibly restarting it during the cooling and recovery period can easily cause mechanical jamming or insulation breakdown. Specifically, the mechanical protection time threshold should be set according to the minimum operating interval specified by the transformer manufacturer, typically ranging from 60 to 300 seconds. Within this time window, regardless of how much the grid voltage exceeds the limit or how the cost function is calculated, the system forcibly blocks all tap change commands, relying entirely on the wide-range adjustment capability of the cascaded step-down device to maintain power supply quality. Only when the interval time meets the safety requirements is a new tap change command allowed to penetrate and be issued.

[0049] By combining the aforementioned cost function with a dual anti-jitter interception mechanism, this embodiment establishes a defensive decision-making system at the algorithm level that balances energy efficiency and lifespan. Compared to the simple logic in existing technologies that relies solely on voltage upper and lower limit thresholds for triggering, the technical solution in this embodiment can not only identify "whether voltage regulation should be implemented," but also determine "whether voltage regulation is worthwhile now" and "whether voltage regulation is safe now." Especially when facing the complex and variable voltage conditions at the end of the distribution network, this mechanism significantly reduces the number of ineffective transformer operations. Actual measurements show that, under the same voltage quality indicators, the daily average number of transformer operations using this embodiment can be reduced by more than 60%, effectively solving the industry pain point of easily damaged mechanical components in joint voltage regulation systems.

[0050] Furthermore, to address the voltage step surge caused by the mechanical shifting of on-load tap-changing transformers, a transient compensation and seamless recovery scheme based on low-level hardware operation is provided. In response to the detection of a step change in the actual input voltage, the voltage closed-loop feedback control is suspended, and a transient feedforward direct control mode is entered. The feedforward duty cycle is calculated based on the actual input voltage after the step and the target output voltage reference value, and the drive signal of the cascaded step-down device is updated. Specifically, this process is an ultra-fast response flow completed within a microsecond-level control cycle. Its core lies in breaking the time lag limitation of traditional closed-loop regulation, achieving real-time suppression of disturbances by sinking the algorithm down to the hardware driver layer.

[0051] In step S501, within the microsecond-level control cycle during which a step change in the actual input voltage is detected, the conventional closed-loop feedback network with integral hysteresis is momentarily suspended. When the main control unit detects through the intermediate voltage sensor that the rate of change of the actual input voltage exceeds the preset step judgment threshold, it confirms that the transformer tap changer has completed physical switching. At this time, if the conventional PI (proportional-integral) closed-loop controller continues to operate, the sudden change in input voltage will cause the feedback error to surge instantaneously, and the integrator will accumulate a large amount of saturation value in a very short time. This integral saturation effect takes several or even dozens of control cycles to subside after the step change ends, during which the controller output is still in an extreme state, inevitably leading to severe overshoot or deep drop in the load-side output voltage. Therefore, the strategy adopted in this embodiment is to immediately freeze the integral term or forcibly clamp the integrator output to near the current steady-state operating point, thereby cutting off the error accumulation path and preventing the controller from entering the nonlinear saturation region. It should be understood that the suspension operation is not limited to freezing the integral; in other embodiments, it can also be to instantaneously set the output weight of the closed-loop controller to zero or reset the controller state to a preset safe initial value, as long as the negative impact of the integral hysteresis effect on the transient process can be blocked.

[0052] Step S502: Directly acquire the transient actual input voltage and recalculate the feedforward duty cycle using the transient feedforward formula. While suspending the closed loop, the system performs open-loop feedforward calculation using the latest sampled true value of the actual input voltage. As a preferred implementation, the formula for calculating the feedforward duty cycle is: In the formula, For feedforward duty cycle, The target output voltage reference value, This represents the actual input voltage of the cascaded buck converter, acquired in real time after the step change. The formula is derived based on the volt-second balance principle of the Buck converter, enabling the direct establishment of a static mapping relationship between the input / output voltage ratio and duty cycle under ideal conditions. It can predict the control quantity required to maintain the target output without relying on error feedback. Compared to the passive mechanism of closed-loop regulation, which requires waiting for errors to occur before correction, feedforward control is an active disturbance cancellation, theoretically eliminating regulation delay. It should be noted that the above formula is an ideal model. In practical engineering, inductor current ripple compensation terms or device voltage drop correction terms can be superimposed to further improve the static accuracy of the feedforward calculation. However, these modifications do not change the fundamental nature of feedforward direct control as a transient-dominant mechanism.

[0053] Step S503 involves converting the feedforward duty cycle into a corresponding timer count value and writing it into the comparison register of the pulse width modulation module to update the drive signal within one pulse width modulation switching cycle. This step is a key hardware-bound feature that distinguishes this embodiment from a pure software algorithm. After calculating the feedforward duty cycle, the main control unit does not simply store it in a memory variable. Instead, it directly accesses the enhanced pulse width modulation (EPWM) peripheral module inside the microcontroller through the underlying driver and writes the converted count value into a shadow register or active comparison register. Since the EPWM module typically supports a shadow loading mechanism, the newly written count value will be automatically loaded and effective when the counter in the current PWM cycle returns to zero or the cycle matches, thus ensuring that the next output PWM pulse immediately adopts the new duty cycle. This direct manipulation of hardware registers bypasses the time overhead caused by operating system scheduling, communication protocol stacks, and complex control algorithm iterations, compressing the delay of the entire compensation action to within one switching cycle (e.g., only 50μs at a 20kHz switching frequency). In contrast, traditional closed-loop regulation is limited by the pipeline time of sample-and-hold, digital filtering, and PI calculation, often requiring 3 to 5 switching cycles to effectively respond to a step change. Voltage runaway during this period is the root cause of output waveform distortion. This embodiment eliminates control delay at the physical level by embedding the control algorithm into real-time hardware actions, achieving precise masking of shift shocks.

[0054] In response to stable operation of the actual input voltage, the voltage closed-loop feedback control is restored. Although transient feedforward direct control can achieve extremely fast response, it is essentially an open-loop control and cannot eliminate steady-state deviations caused by device parameter drift, dead-zone effects, and load changes. Therefore, when the system determines that the fluctuation rate of the actual input voltage has fallen below the steady-state threshold and the duration exceeds the preset stability confirmation window, it must smoothly exit the transient feedforward direct control mode and restore the voltage dual closed-loop feedback control network, which includes an outer voltage loop and an inner current loop.

[0055] In practice, the recovery process is not a simple switching operation, but rather employs a smooth transition strategy to avoid secondary disturbances. For example, the feedforward duty cycle at the moment of feedforward exit can be used as the initial bias of the closed-loop controller output. This bias is then linearly decayed over several control cycles until it reaches zero, while the output authority of the closed-loop controller is gradually released. Alternatively, a weighted fusion approach can be used, where the feedforward output and closed-loop output are dynamically weighted during the transition period, with the weighting coefficients smoothly transitioning from feedforward dominance to closed-loop dominance. The recovered voltage outer loop primarily eliminates the system's steady-state error, utilizing its high-gain characteristics to ensure long-term accurate tracking of the target output voltage reference value by the load-side output voltage. The current inner loop provides system damping and current-limiting protection, suppressing the resonant peak of the LC filter network through rapid closed-loop adjustment of the inductor current, and achieving microsecond-level current limiting during load surges or short-circuit faults, ensuring that the cascaded step-down device continuously outputs a rated voltage that meets the target output voltage reference value to the load. Through this seamless connection between transient feedforward and steady-state dual closed loop, this embodiment not only ensures the dynamic performance at the moment of gear shifting, but also maintains the static accuracy and safety of long-term operation, perfectly solving the problem of timing mismatch between mechanical action and power electronic control in hybrid voltage regulation systems.

[0056] This embodiment provides a combined voltage regulation system for a distribution network transformer and a cascaded step-down device. This system serves as the physical execution carrier of the aforementioned combined voltage regulation method, achieving deep decoupling between macroscopic discrete voltage regulation and microscopic continuous voltage regulation through the coordinated cooperation of the hardware architecture. Specifically, the combined voltage regulation system includes a main power unit, a state detection network, a main control unit, and a bypass protection unit.

[0057] The main power unit includes an on-load tap-changing transformer, a cascaded step-down device, and an output filter network connected in series. Figure 1 In the illustrated topology, electrical energy flows in from the distribution network, undergoes coarse adjustment by an on-load tap-changing transformer, then enters a cascaded step-down device for fine adjustment, and finally passes through an output filter network to filter out high-frequency switching ripple before being delivered to the load. The cascaded step-down device can employ a Buck converter topology, and the output filter network can use a CLC-type filter structure to balance size and filtering performance. It should be understood that although the diagram shows a vertically series energy transmission path, in other implementations, the physical layout of each power module can be adaptively adjusted according to the cabinet space, as long as the electrical series cascade relationship is maintained. This hardware cascade structure provides the physical basis for the "micro-level voltage regulation priority, macro-level tap adjustment as a safety net" control strategy in the aforementioned embodiments, enabling the cascaded step-down device to independently absorb most voltage fluctuations, while the on-load tap-changing transformer only needs to perform low-frequency tap adjustments when necessary.

[0058] The condition monitoring network includes a grid voltage sensor for monitoring the background voltage of the distribution network, an intermediate voltage sensor for acquiring the actual input voltage of the cascaded step-down device, a load-side sensor for acquiring the output voltage and load current on the load side, and an inductor current sensor for acquiring the inductor current inside the cascaded step-down device. These four types of sensors constitute the system's full-dimensional sensing system. It is particularly important to emphasize that the intermediate voltage sensor is installed between the secondary output terminal of the on-load tap-changing transformer and the input terminal of the cascaded step-down device. This specific arrangement is crucial for achieving the aforementioned transient feedforward direct control. Only by directly sampling at the physical input port of the cascaded step-down device can the true voltage step value generated by the instantaneous action of the mechanical contacts of the on-load tap-changing transformer be accurately captured, thus providing a delay-free data source for feedforward calculations. If this sensor is moved to the primary side of the transformer or only relies on communication to obtain tap position signals, feedforward compensation will fail due to changes in line impedance or communication delays. In addition, the inductive current sensor uses a high-frequency Hall element or a shunt, and its sampling bandwidth needs to cover the switching frequency of the cascaded step-down device in order to provide the main control unit with accurate current inner loop feedback and ripple observation data.

[0059] The main control unit communicates with the main power unit and the state detection network, and is configured to execute the joint voltage regulation method described above. In specific implementations, the main control unit is typically composed of a high-performance digital signal processor (DSP) or a field-programmable gate array (FPGA), or adopts a dual-core heterogeneous architecture of DSP+FPGA. The FPGA is responsible for the underlying high-speed signal acquisition, PWM generation, and transient feedforward direct control logic, ensuring that register refresh is completed within one switching cycle; the DSP is responsible for running complex algorithms such as global comprehensive loss model, cost function optimization, and anti-jitter interception. This hierarchical configuration of computing power ensures that the system can respond to microsecond-level transient disturbances and handle millisecond-level steady-state optimizations, avoiding control timing conflicts caused by task overload of a single processor. The main control unit reads the four analog signals from the state detection network in real time through a high-speed AD interface, and issues shifting commands to the on-load tap changer via GPIO or the communication bus, while simultaneously outputting high-precision PWM drive signals to the cascaded step-down device, realizing a closed loop between control logic and power execution.

[0060] The bypass protection unit is connected to the main control unit and is configured to block the drive pulse and close the bypass switch when a fault is detected in the internal components of the cascaded step-down device or an extreme voltage drop in the grid. This allows the on-load tap-changing transformer to bypass the cascaded step-down device and directly supply power to the load. Specifically, the bypass protection unit includes a fast mechanical contactor or a thyristor solid-state switch connected in parallel across the cascaded step-down device. When the main control unit detects a serious fault such as overcurrent, overtemperature, or DC bus undervoltage, it will forcibly shut down all power transistor drive signals of the cascaded step-down device within microseconds and then issue a bypass closing command. Because the closing action of the bypass switch and the shut-off action of the power electronic devices are strictly interlocked in time, circulating current short circuits can be effectively prevented. After the bypass is activated, the system automatically degrades to the traditional on-load tap-changing transformer-only power supply mode. Although the fine voltage regulation capability is temporarily lost, the power supply continuity of the distribution area is guaranteed to the greatest extent, avoiding large-scale power outages caused by power electronic equipment failures. This redundancy design significantly improves the engineering practicality and safety margin of the combined voltage regulation system.

[0061] Through the synergy of the aforementioned hardware architecture, this embodiment provides reliable structural support for the joint voltage regulation method. The precise layout of the state detection network solves the timeliness problem of transient sensing, the heterogeneous computing power allocation of the main control unit meets the needs of multi-timescale control, and the bypass protection unit provides physical protection for the system's safety baseline. The functional definitions and connections of each hardware module together constitute the necessary material conditions for achieving global efficiency optimization and seamless transient response.

[0062] This embodiment provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the combined voltage regulation method as described above.

[0063] Specifically, the term "computer-readable storage medium" as used in this application is intended to encompass all tangible media capable of carrying, storing, transmitting, or propagating sequences of instructions for use by an instruction execution system, apparatus, or device. As a preferred embodiment, the storage medium may be a non-volatile memory, such as read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, optical discs (such as CD-ROMs and DVDs), magnetic disks (such as floppy disks and hard disks), or magnetic tape. In other embodiments, the storage medium may also be a volatile memory, such as random access memory (RAM), as long as it can retain program code during device operation. It should be understood that the above examples are merely exemplary and not an exhaustive limitation on the types of storage media; any physical carrier capable of storing data in electronic, magnetic, optical, electromagnetic, infrared, or semiconductor manner falls within the scope of protection of this application.

[0064] When the processor in the main control unit reads and executes the aforementioned computer program from the computer-readable storage medium, it can transform the abstract control logic in the foregoing embodiments into specific hardware drive signals. For example, the processor executes the global integrated loss modeling module in the program, calls the underlying AD sampling interface to obtain the grid voltage and load current, and constructs an evaluation function in memory that includes transformer losses and cascaded step-down device losses; executes the cost function optimization module, traverses candidate gears to calculate expected losses and determines the target gear by combining wear penalty weights; executes the anti-jitter interception module, compares the voltage change trend with the mechanical protection time threshold to determine whether to issue a gear shift command; and executes the transient feedforward direct control module, suspending the closed-loop integrator and directly refreshing the PWM comparator register within a microsecond-level period when a step change in the input voltage is detected. In this way, the software program stored in the medium becomes a bridge connecting the control algorithm and the power hardware, enabling the joint voltage regulation method to be distributed, sold, or updated via network download as an independent product, independent of a specific development environment.

[0065] The technical advantage of this embodiment lies in that, by solidifying the combined voltage regulation method into a computer program product, it not only ensures the consistency and portability of the control strategy across different batches and models of main control units, but more importantly, it provides a direct basis for intellectual property protection against pure software objects or embedded firmware. In practical engineering applications, even if the infringing party does not manufacture the complete combined voltage regulation system hardware, but only sells or offers to sell storage media containing the program, or provides download services for the program via the network, it still falls within the scope of protection of this application, thereby effectively making up for the potential blind spots in rights protection that may exist if only the method or system hardware is protected.

[0066] To more intuitively verify the technical effect of the combined voltage regulation method provided in this application in actual engineering, this embodiment takes a typical photovoltaic grid connection scenario as an example for detailed explanation. It should be understood that the specific values ​​and operating conditions below are merely illustrative examples, intended to clarify the operating mechanism and beneficial effects of the technical solution, and are not intended to limit the scope of protection of this application. In this embodiment, the application scenario is set as a 10kV / 0.4kV distribution grid area connected to a large-capacity distributed photovoltaic power generation system, with mixed residential and commercial loads. The system parameters are configured as follows: the target output voltage reference value V_ref is set to 220V, the voltage dead zone threshold e is set to 3V, the equipment wear penalty weight coefficient W is set to 100W / level, the switching frequency of the cascaded step-down device is 20kHz, and the control cycle of the main control unit is 50μs.

[0067] During peak photovoltaic (PV) output at midday, the voltage in the distribution area often rises significantly due to reverse power flow. Assume that the rise in grid background voltage causes the input voltage of the cascaded step-down converter to increase to 280V, while the load side needs to maintain a 220V output. According to the "micro-level voltage regulation priority" logic established in Example 1, the system first uses the cascaded step-down converter to absorb the excess 60V voltage difference by adjusting the PWM duty cycle. Calculations show that the duty cycle required to maintain rated output at this time is approximately 0.79 (considering device voltage drop correction). This value is within the efficient operating range of the cascaded step-down converter, and the current tap position of the on-load tap changer does not cause the total system loss to deviate from the optimal point. Therefore, the global comprehensive loss model determines that macro-level tap adjustment is unnecessary, and the on-load tap changer remains stationary. This process fully demonstrates the power electronic converter's rapid absorption capability for high-frequency, large-amplitude voltage fluctuations, avoiding the ineffective response of the on-load tap changer to short-term PV fluctuations.

[0068] However, as photovoltaic output further increases or grid voltage continues to rise, the input voltage of the cascaded step-down device may continue to rise above 320V, causing the duty cycle required to maintain 220V output to drop below 0.6. When this low duty cycle state continues for more than a preset time threshold (e.g., 5 minutes), the system determines that the current operating point has deviated from the optimal efficiency range, and the conduction loss and inductor ripple loss of the cascaded step-down device increase significantly. At this time, the global comprehensive loss model and cost function optimization mechanism are activated. The system traverses the candidate taps of the on-load tap-changing transformer and calculates that if the transformer is raised by one tap, although its secondary no-load output voltage decreases, the predicted input voltage after combining the line voltage drop will drop back to around 260V, and the corresponding duty cycle of the cascaded step-down device can be raised back to around 0.85. Although upgrading the transformer's tap level slightly increases its own copper losses, the reduction in switching and inductance losses due to the lower input voltage of the cascaded step-down device far outweighs the increased losses of the transformer. Furthermore, after deducting the wear penalty of W=100W / taper, the overall cost J(k) after upgrading is still significantly lower than the cost of maintaining the current tap level. Therefore, after passing dual anti-jitter interception verification, the system issues an upgrade command to the on-load tap-changing transformer. Actual measurement data shows that after this tap level optimization adjustment, the overall system operating efficiency improved by approximately 2%, verifying the effectiveness of global energy efficiency collaborative optimization.

[0069] At the instant the on-load tap-changing transformer performs its upshift, the actual input voltage of the cascaded step-down device experiences a step change of approximately 10V due to the switching of mechanical contacts. The transient feedforward direct control mechanism plays a crucial role in addressing this transient process. The main control unit captures this step signal within microseconds using an intermediate voltage sensor and immediately suspends the integrator in the voltage closed-loop feedback control to prevent overshoot caused by integral saturation. Subsequently, based on the actual input voltage after the step change and the target output voltage reference value of 220V, the system recalculates the duty cycle using the feedforward formula and writes the new count value to the EPWM comparator register within 20μs, ensuring the drive signal is updated within one switching cycle. Thanks to this ultra-fast compensation at the underlying hardware level, the maximum fluctuation of the load-side output voltage during the entire shift process is strictly limited to within 1V, achieving a smooth transition that is imperceptible to the user. In contrast, if the transient feedforward direct control mechanism of this application is not adopted, and only traditional PI closed-loop regulation is used, due to integral lag and system inertia, the same 10V input step will cause the output voltage to have a transient overshoot of more than 15V, and the regulation stabilization time is as long as several milliseconds, which seriously affects the power supply quality of sensitive loads.

[0070] Furthermore, long-term operational statistics show that after adopting the combined voltage regulation method provided in this application, the average daily number of operations of the on-load tap-changing transformer in this distribution area has been reduced by more than 60% compared to the traditional control strategy based on upper and lower voltage thresholds. This is mainly attributed to the introduction of wear penalty weights in the cost function and the filtering effect of the dual anti-jitter interception mechanism, which ensures that the transformer only performs actions when it truly achieves significant energy efficiency benefits and safety conditions are met, while the vast majority of voltage regulation tasks are handled by cascaded step-down devices that have no mechanical wear. This hierarchical collaboration not only significantly extends the maintenance cycle and service life of the on-load tap-changing transformer, but also reduces the overall system loss through dynamic optimization of the operating point, achieving a dual improvement in economic benefits and equipment safety. It should be understood that the specific parameters in the above embodiments (such as W=100, e=3V, 20μs response time, etc.) are all preferred values ​​set according to specific equipment specifications and operating conditions. In practical applications, those skilled in the art can make adaptive adjustments based on transformer capacity, power electronic device characteristics, and power grid quality standards, and these adjustments should not deviate from the technical concept and protection scope of this application.

[0071] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method and system for combined voltage regulation of a distribution network transformer and a cascaded step-down device, characterized in that, include: In response to the load-side output voltage deviating from the target output voltage reference value, a global comprehensive loss model is constructed, which includes the losses of the on-load tap-changing transformer and the losses of the cascaded step-down device. Based on the global comprehensive loss model, the expected total system loss corresponding to multiple candidate taps of the on-load tap-changing transformer is evaluated, and a cost function is constructed by combining the equipment wear penalty weight to determine the target tap. Under the condition that the preset anti-shake interception conditions are met, a control command to switch to the target gear is sent to the on-load tap-changing transformer. During the tap-changing process of the on-load tap-changing transformer, the actual input voltage of the cascaded step-down device is monitored; In response to the detection of a step change in the actual input voltage, the voltage closed-loop feedback control is suspended, and the transient feedforward direct control mode is switched to. The feedforward duty cycle is calculated based on the actual input voltage after the step change and the target output voltage reference value, and the drive signal of the cascaded buck converter is updated. In response to the stable operation of the actual input voltage, the voltage closed-loop feedback control is restored.

2. The combined voltage regulation method according to claim 1, characterized in that, The construction of the global comprehensive loss model, which includes the losses of on-load tap-changing transformers and cascaded step-down devices, includes: Construct a comprehensive loss evaluation function that covers the losses of on-load tap-changing transformers, conduction losses of cascaded step-down devices, switching losses, and filter inductor losses; The evaluation of the expected total system losses corresponding to multiple candidate tap positions of the on-load tap-changing transformer includes: For each candidate tap, the predicted input voltage of the cascaded step-down device when switching to that candidate tap is calculated based on the grid-side background voltage and load current. Substituting the predicted input voltage into the comprehensive loss evaluation function, the expected total system loss corresponding to maintaining rated output when switching to the candidate gear is calculated.

3. The combined voltage regulation method according to claim 2, characterized in that, The formula for calculating the predicted input voltage is as follows: In the formula, To switch to the 1st Predicted input voltage at gear position The background voltage is collected in real time from the power grid side. The on-load tap-changing transformer is the first The rated no-load output voltage of the gear position. For load current, For the first Equivalent short-circuit impedance at the specified gear position.

4. The combined voltage regulation method according to claim 1, characterized in that, The cost function constructed by combining equipment wear penalty weights includes: The expected total system loss is combined with a penalty term representing the equipment wear cost caused by mechanical shifting to construct the cost function for evaluating the overall economic efficiency of candidate gears, so as to avoid frequent switching between similar optimal solutions due to grid voltage fluctuations. The step of issuing a control command to the on-load tap-changing transformer to switch to the target tap, under the condition that the preset anti-shake interception conditions are met, includes: Execute the first level of security interception judgment: verify the difference between the target level and the current level and the current grid voltage change trend. If it is determined that the difference between the target level and the current level is one level and the current grid voltage change trend is reversing, then the control command is temporarily suspended. The second layer of safety interception judgment is executed: the interval time since the last gear shift action is read, and if the interval time is less than the preset mechanical protection time threshold, the control command is blocked; if it is greater than or equal to the mechanical protection time threshold, the control command is officially issued.

5. The combined voltage regulation method according to claim 4, characterized in that, The formula for calculating the cost function is as follows: In the formula, For the first The comprehensive cost function value of each candidate gear position. To switch to the 1st The predicted total system loss at the gear position. This is the weighting coefficient for equipment wear penalty. This refers to the tap number of the on-load tap-changing transformer currently under consideration. The current actual tap position of the on-load tap-changing transformer; the control command is issued to the target tap position corresponding to the tap position number that minimizes the value of the comprehensive cost function.

6. The combined voltage regulation method according to claim 1, characterized in that, In response to detecting a step change in the actual input voltage, the voltage closed-loop feedback control is suspended, and a transient feedforward direct control mode is entered. The feedforward duty cycle is calculated based on the actual input voltage after the step change and the target output voltage reference value, and the drive signal of the cascaded buck converter is updated, including: Within the microsecond-level control cycle that detects a step change in the actual input voltage, the conventional closed-loop feedback network with integral hysteresis is instantaneously suspended. The actual input voltage in the transient state is directly acquired, and the feedforward duty cycle is recalculated using the transient feedforward formula; The feedforward duty cycle is converted into a corresponding timer count value and written into the comparison register of the pulse width modulation module so as to update the drive signal within one pulse width modulation switching cycle.

7. The combined voltage regulation method according to claim 6, characterized in that, The response to the stable operation of the actual input voltage, restoring the voltage closed-loop feedback control, includes: Once the actual input voltage step is determined to have ended and the system is operating stably at the voltage corresponding to the target level, the system smoothly exits the transient feedforward direct control mode and restores the voltage dual closed-loop feedback control network, which includes an outer voltage loop and an inner current loop. The outer voltage loop is used to eliminate system steady-state errors, and the inner current loop is used to provide system damping and current limiting protection to ensure that the cascaded step-down device continuously outputs a rated voltage to the load that meets the target output voltage reference value.

8. The combined voltage regulation method according to claim 1, characterized in that, In response to the load-side output voltage deviating from the target output voltage reference value, a global comprehensive loss model is constructed, including the losses of the on-load tap-changing transformer and the cascaded step-down device, comprising: Set the target output voltage reference value and the allowable voltage dead zone threshold; When continuous When the absolute value of the difference between the load-side output voltage and the target output voltage reference value is detected to be greater than the voltage dead zone threshold within a power frequency cycle, the construction of the global comprehensive loss model and the determination of the target level are triggered. It is a positive integer; When the absolute value of the difference is determined to be less than or equal to the voltage dead zone threshold, it is determined that the current state is in the optimal steady state and the current operating state is maintained.

9. A combined voltage regulation system for a distribution network transformer and a cascaded step-down device, characterized in that, include: The main power unit includes an on-load tap-changing transformer, a cascaded step-down device, and an output filter network connected in series. The condition detection network includes a grid voltage sensor for monitoring the background voltage of the distribution network, an intermediate voltage sensor for acquiring the actual input voltage of the cascaded step-down device, a load-side sensor for acquiring the load-side output voltage and load current, and an inductor current sensor for acquiring the inductor current inside the cascaded step-down device. The main control unit, which is communicatively connected to the main power unit and the status detection network, is configured to execute the joint voltage regulation method as described in any one of claims 1 to 8; The bypass protection unit, connected to the main control unit, is configured to block the drive pulse and close the bypass switch when a fault is detected in the internal components of the cascaded step-down device or an extreme drop in the grid voltage, so that the on-load tap-changing transformer can bypass the cascaded step-down device and directly supply power to the load.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the combined voltage regulation method as described in any one of claims 1 to 8.

Images