Local side energy conversion method and device for converting alternating current into direct current and boosting to kilovolt

Abstract

The invention discloses a local side energy conversion method and device for converting alternating current into direct current and boosting to kilovolt, and belongs to the technical field of voltage conversion, and the method specifically comprises the steps: collecting an input alternating current signal, and generating a first pulse alternating current signal which is in phase locking with an alternating current zero crossing point according to the collected input alternating current signal, inputting the first pulse alternating current signal into a sectional type capacitance voltage-multiplying network formed by cascading a plurality of boosting units, setting a resonance branch at an intermediate node of the sectional type capacitance voltage-multiplying network, and generating capacitance-inductance resonance within a preset frequency range to obtain a second pulse voltage signal, performing direct current collection on the second pulsating voltage signal, and adjusting parameters of the first pulse alternating current signal according to voltage state feedback of each boosting unit to generate stable kilovolt-level direct current voltage; according to the invention, the size and deployment limitations are significantly reduced, the adaptability to complex input conditions is enhanced, and the practicability and application range of local-side high-voltage direct-current energy supply are improved.

Description

Technical Field

[0001] This invention belongs to the field of voltage conversion technology, specifically a local-end energy conversion method and device for AC to DC voltage boosting to kilovolts. Background Technology

[0002] Existing AC-to-DC high-voltage power supply solutions mainly rely on centralized power conversion structures, obtaining high-voltage DC output through power frequency or medium-to-high frequency transformers, rectification, and filtering. These solutions typically require large magnetic components, complex insulation structures, and strict safety clearances, resulting in a large overall system size, poor deployment flexibility, and difficulty in meeting the high-voltage DC power supply needs of terminal sides, central office sides, or space-constrained environments. Furthermore, centralized boosting modes easily introduce additional losses during long-distance transmission, further increasing the complexity of the system architecture.

[0003] With the development of distributed power electronics technology, boost schemes based on voltage multiplier rectification, capacitor cascading, and switch control have emerged to replace traditional transformer boost structures. However, existing voltage multiplier technologies mostly adopt a fixed number of stages and static operation mode. Their voltage build-up process lacks fine-grained timing correlation with the input AC waveform. In local power conversion scenarios, the waveform characteristics, frequency stability, and load changes of the input AC signal are often uncertain. Traditional boost and rectification methods that operate at fixed frequency or fixed parameters cannot effectively coordinate the relationship between voltage build-up, energy transfer, and output regulation. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention proposes a local-end energy conversion method and apparatus for AC-to-DC boosting to kilovolts. This method enables coordinated control of the timing characteristics, charge migration process, and output collection mechanism of AC signals without the need for large magnetic components, thereby achieving kilovolt-level DC energy conversion under local-end conditions.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for AC-to-DC voltage boosting to kilovolts at the local power supply end includes: The system acquires the input AC signal and generates a first pulse AC signal that is phase-locked with the zero-crossing point of the AC signal based on the acquired input AC signal. The first pulsed AC signal is input into a preset segmented capacitor multiplier network to obtain a second pulsating voltage signal. The segmented capacitor multiplier network is composed of multiple cascaded boost units, with a resonant branch set in the middle node to generate capacitor-inductor resonance within a preset frequency range. The second pulsating voltage signal is DC collected, and the parameters of the first pulse AC signal are adjusted according to the voltage status feedback of each boost unit to generate a stable kilovolt-level DC voltage.

[0006] Specifically, the step of acquiring the input AC signal and generating a first pulse AC signal that is phase-locked with the zero-crossing point of the AC signal based on the acquired input AC signal includes: Within a continuous sampling period, the input AC signal is sampled at multiple time points to obtain a sampling data sequence that reflects the periodic changes of the AC signal. Perform sign change analysis on the sampled data sequence to identify time segments in the sampled data where the voltage polarity changes, and form a set of zero-crossing candidate segments; The zero-crossing candidate segment set is subjected to time consistency screening to determine the zero-crossing reference time corresponding to the current sampling period, and a time identifier corresponding to the zero-crossing reference time is established; A continuous phase evolution relationship is constructed based on the time identifier, and a trigger time sequence is generated according to a preset phase update rule; Based on the trigger time sequence, pulse processing is performed within the corresponding time interval of the input AC signal to generate a first pulse AC signal that maintains a timing correspondence with the zero-crossing reference time.

[0007] Specifically, based on the trigger time sequence, pulse processing is performed within the corresponding time interval of the input AC signal to generate a first pulse AC signal that maintains a timing correspondence with the zero-crossing reference time, including: The trigger time sequence is mapped to a set of pulse windows aligned with the time axis of the input AC signal, and a window identifier associated with the zero-crossing reference time is written for each pulse window; The set of pulse windows is subjected to window validity verification, and windows that overlap in time or conflict in sequence with adjacent windows are removed according to preset order constraint rules to obtain a set of pulse windows that have been constrained. Using the constrained pulse window set as the boundary, the input AC signal is segmented and sampled and segmented to form a pulse segment sequence corresponding to each pulse window, and the time interval outside the pulse window is set to a preset potential state. The pulse segment sequence is time-sequentially spliced ​​according to the window identifier, and a first pulse AC signal that maintains a time-sequential correspondence with the zero-crossing reference time is output.

[0008] Specifically, the first pulsed AC signal is input into a preset segmented capacitor multiplier network to obtain a second pulsating voltage signal, including: Based on the first pulse AC signal, a charging time slot sequence and a transfer time slot sequence corresponding to the first pulse AC signal are set for multiple boost units, and a time slot identifier is assigned to each boost unit; According to the time slot identifier, the segmented capacitor multiplier network is controlled to enter a phased charging state, so that each boost unit completes charge writing within the corresponding charging time slot and generates a charge state sequence. Based on the charge state sequence, cross-cell charge transfer scheduling is performed. During the transfer time slot, adjacent boost cells are subjected to charge transfer and potential superposition processing to form a node potential sequence, and the node potential sequence is written into the node time stamp queue. In the segmented capacitor multiplier network, a resonant intervention period corresponding to the node time stamp queue is selected at the intermediate node, and a resonant branch is introduced to participate in charge exchange during the resonant intervention period to generate a resonant participation identifier sequence. Within a preset frequency range, the node potential sequence is updated in time according to the resonant participation identifier sequence, and a second pulsating voltage signal corresponding to the node potential sequence is output.

[0009] Specifically, based on the charge state sequence, cross-cell charge transfer scheduling is performed. During the transfer time slot, adjacent boost cells undergo charge transfer and potential superposition processing to form a node potential sequence, which is then written into the node time stamp queue. This includes: The charge state sequence is analyzed at the cell level, and a migration candidate set of adjacent boost cell pairs is generated according to a preset migration rule. The source cell, target cell and transfer time slot are recorded for each migration candidate. The migration candidate set is subjected to conflict resolution processing. According to the preset mutual exclusion constraints, migration candidates that share the same boost unit or share the same transfer time slot are screened out and rearranged to obtain the migration execution sequence. Within each transfer time slot corresponding to the migration execution sequence, charge migration and potential superposition operations are sequentially performed on the source unit and the target unit, and the instantaneous potential of the adjacent boost unit connection node is collected after each operation to form a node potential sequence arranged in time. The node potential sequence is appended with a time tag corresponding to the transfer time slot, and written into the node time tag queue in the order of the time tags.

[0010] Specifically, in the segmented capacitor multiplier network, a resonant intervention period corresponding to the node time-stamped queue is selected at the intermediate node, and a resonant branch is introduced to participate in charge exchange during the resonant intervention period to generate a resonant participation identifier sequence, including: The node time stamp queue is divided into time slices, the node potential change segments corresponding to each time stamp are extracted, and each node potential change segment is converted into a set of node change events. The set of node change events is filtered to determine target events that meet the triggering conditions according to preset resonance triggering rules, and each target event is assigned a corresponding resonance intervention time period identifier to form a set of resonance intervention time periods. Within each resonance intervention period defined by the set of resonance intervention periods, resonance branch intervention operation is performed, and the state of the charge exchange process of the intermediate node during the resonance intervention period is recorded to generate resonance participation record corresponding to each resonance intervention period identifier. The resonance participation record is encoded and a resonance participation identifier sequence is output in chronological order. The resonance participation identifier sequence is used to characterize the participation status of the resonance branch during each resonance intervention period.

[0011] Specifically, the step of updating the node potential sequence in time according to the resonant participation identifier sequence within a preset frequency range and outputting a second pulsating voltage signal corresponding to the node potential sequence includes: The resonance participation identifier sequence is time-aligned and mapped to the time label of the node potential sequence to obtain the resonance mapping sequence. Based on the resonant mapping sequence, the node potential sequence is divided into resonant update segments and non-resonant update segments, and update rule sets corresponding to each segment are generated respectively. Within the preset frequency range, each resonant update segment and each non-resonant update segment are sequentially updated according to the set of update rules to obtain the updated node potential sequence. The updated node potential sequence is subjected to timing shaping, and a second pulsating voltage signal corresponding to the updated node potential sequence is output.

[0012] Specifically, the updated node potential sequence is subjected to time-series shaping, and a second pulsating voltage signal corresponding to the updated node potential sequence is output, including: The updated node potential sequence is segmented according to time labels, potential change segments between adjacent time labels are extracted to form a segment set, and a segment identifier is written for each segment. The segment set is subjected to boundary consistency processing. Segments with gaps are filled and segments with overlaps are trimmed according to preset boundary rules to obtain a segment set with boundary processing. The set of segments after boundary processing is subjected to sequence arrangement processing. The segment identifiers are rearranged according to the preset time arrangement rules, and the rearranged segments are spliced ​​together in the corresponding order to form a time-shaped node potential sequence. Based on the time-shaped node potential sequence, a second pulsating voltage signal corresponding to the time-shaped node potential sequence is output.

[0013] Specifically, the second pulsating voltage signal is DC collected, and the parameters of the first pulsed AC signal are adjusted according to the voltage status feedback of each boost unit to generate a stable kilovolt-level DC voltage, including: The second pulsating voltage signal is unidirectionally processed according to a preset rectification timing, and the unidirectional signal is written into the DC converging sequence. The DC collection sequence is sampled hierarchically to obtain a voltage state set corresponding to each boost unit, and the voltage state set is written into a state time stamp queue. A parameter update sequence is constructed based on the state time stamp queue. The parameter update sequence includes update instructions for the pulse width parameter, pulse interval parameter, and pulse cluster length parameter of the first pulse AC signal. The parameters of the first pulse AC signal are iteratively adjusted according to the parameter update sequence, and a kilovolt-level DC voltage corresponding to the DC collection sequence is output.

[0014] An AC-to-DC voltage boosting device for local-end energy conversion, used to implement the aforementioned AC-to-DC voltage boosting method for local-end energy conversion, includes: a first signal generation module, a second signal generation module, and a kilovolt voltage output module; The first signal generation module is used to acquire the input AC signal and generate a first pulse AC signal that is phase-locked with the zero-crossing point of the AC signal based on the acquired input AC signal. The second signal generation module is used to input the first pulse AC signal into a preset segmented capacitor voltage multiplier network to obtain a second pulsating voltage signal. The segmented capacitor voltage multiplier network is composed of multiple cascaded boost units, and a resonant branch is set in the middle node to generate capacitor-inductor resonance within a preset frequency range. The kilovolt voltage output module is used to collect the second pulsating voltage signal into DC and adjust the parameters of the first pulse AC signal according to the voltage status feedback of each boost unit to generate a stable kilovolt-level DC voltage.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention proposes a method and apparatus for AC-to-DC voltage boosting to kilovolts at the central office. By timing the input AC signal, a pulsed excitation signal associated with the AC zero-crossing point is constructed. Charge writing, migration, superposition, and resonance intervention are completed in a segmented capacitor multiplier network according to a controlled timing sequence. Finally, a kilovolt-level DC output is formed by combining DC collection and feedback regulation. This method unifies the time characteristics of the AC signal, the charge migration path, and the output regulation process, making the energy conversion process controllable and scalable at the central office. This achieves high-voltage DC output without the need for a power frequency core transformer, significantly reducing size and deployment limitations, enhancing adaptability to complex input conditions, and improving the practicality and application range of high-voltage DC power supply at the central office. Attached Figure Description

[0016] Figure 1 A flowchart of an AC-to-DC boosting method for local-end energy conversion provided by the present invention; Figure 2 This is a schematic diagram of the generation of the first pulse AC signal provided by the present invention; Figure 3 This is a schematic diagram of the generation of the second pulsating voltage signal provided by the present invention; Figure 4 This is a schematic diagram of an AC-to-DC booster power conversion device for kilovolts provided by the present invention. Detailed Implementation

[0017] The present application will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any way. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application. These all fall within the protection scope of the present application.

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] It should be noted that, unless there is a conflict, the various features in the embodiments of this application can be combined with each other, all of which are within the protection scope of this application. Furthermore, although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than the module division in the device or the order in the flowchart. In addition, the terms "first," "second," and "third" used in this application do not limit the data or execution order, but only distinguish identical or similar items with essentially the same function and effect.

[0020] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The term "and / or" as used in this specification includes any and all combinations of one or more of the associated listed items.

[0021] Example 1 Please see Figures 1-3 The present invention provides an embodiment of a local-end energy conversion method for AC-to-DC boosting to kilovolts. This method converts low-voltage AC to kilovolt-level DC without using a power frequency core transformer, and is applied to DC high-voltage aging power supply conversion scenarios. The method includes the following specific steps: Step S1: Acquire the input AC signal and generate a first pulse AC signal that is phase-locked with the zero-crossing point of the AC signal based on the acquired input AC signal.

[0022] like Figure 2 As shown, the specific steps of step S1 are as follows: Step S101: During the continuous sampling period, the input AC signal is sampled at multiple time points to obtain a sampling data sequence that reflects the periodic changes of the AC signal.

[0023] Specifically, the acquisition process is not limited to a single fixed sampling point, but rather selects multiple time positions within the same AC cycle for sampling based on a preset time resolution strategy. This allows the sampled data to cover the rising segment, falling segment, and polarity transition neighborhood of the AC signal on the time axis. The sampled data sequence obtained in this way forms a continuous mapping relationship in time arrangement, enabling the numerical changes between adjacent sampling points to characterize the evolution trend of the input AC signal within the cycle.

[0024] It should be noted that this sampled data sequence is not used directly for voltage calculation, but rather serves as the basic data source for subsequent time series analysis and zero-crossing reference extraction. The use of multiple time points avoids the omission of transient changes in AC waveforms by single-point sampling.

[0025] Step S102: Perform sign change analysis on the sampled data sequence to identify time segments in the sampled data where the voltage polarity changes, and form a set of zero-crossing candidate segments.

[0026] In this embodiment, based on the sampling data sequence obtained in step S101, the polarity changes of the sampling data over time are analyzed by comparing the sign attributes of adjacent sampling data one by one, in order to identify the time segment in which the voltage polarity of the input AC signal changes within the period. Specifically, each sampling point in the sampling data sequence is assigned a positive or negative sign identifier, and the relationship between the sign identifiers of adjacent sampling points is detected along the time sequence. When a sign changes from positive to negative or from negative to positive, the time range covered by the adjacent sampling point is determined to be a candidate segment where the voltage polarity changes. It should be noted that since the sampling data has temporal continuity, the sign change does not occur instantaneously, but is distributed within a finite time width. Therefore, by merging adjacent segments that continuously exhibit sign change characteristics, a set of zero-crossing candidate segments with clear time boundaries is formed.

[0027] Step S103: Perform time consistency screening on the set of zero-crossing candidate segments, determine the zero-crossing reference time corresponding to the current sampling period, and establish a time identifier corresponding to the zero-crossing reference time.

[0028] In this embodiment, for the set of zero-crossing candidate segments formed in step S102, consistency screening is performed according to the time distribution relationship within the sampling period to determine the zero-crossing reference time corresponding to the current sampling period. Specifically, the start and end positions of each zero-crossing candidate segment on the time axis are sorted, and the candidate segments are classified in conjunction with the sampling period boundary. Segments located at the intersection of adjacent sampling periods or those that appear repeatedly are eliminated, and only candidate segments with continuous and reasonable time positions within the current sampling period are retained. Subsequently, a representative time position is selected from the retained candidate segments as the zero-crossing reference time. This time position is used to characterize the polarity conversion point of the AC signal within the sampling period. It should be noted that by performing time consistency screening on the candidate segments, interference from multiple candidate segments caused by noise or sampling jitter can be avoided, thereby making the determined zero-crossing reference time unique within the sampling period, and further establishing a corresponding time identifier for the zero-crossing reference time.

[0029] Step S104: Construct a continuous phase evolution relationship based on the time identifier, and generate a trigger time sequence according to the preset phase update rule.

[0030] In this embodiment, the time marker established in step S103 is used as the phase starting reference to continuously describe the phase change of the input AC signal within the sampling period. Specifically, the time marker is mapped onto the time axis of the sampling period as the initial node for phase evolution, and a continuous correlation between phase and time changes is constructed according to the time progression order within the sampling period, so that any time position can correspond to a unique phase state. Subsequently, according to a preset phase update rule, the phase evolution relationship is discretized, and a corresponding trigger time is generated when the phase progresses to a preset position, thereby forming a trigger time sequence arranged in chronological order. It should be noted that by introducing the time marker into the phase evolution modeling, the trigger time sequence can be kept consistent with the periodic characteristics of the AC signal, avoiding trigger offset caused by sampling period fluctuations.

[0031] Step S105: Based on the trigger time sequence, pulse processing is performed within the corresponding time interval of the input AC signal to generate a first pulse AC signal that maintains a timing correspondence with the zero-crossing reference time.

[0032] The specific steps of step S105 are as follows: Step S1051: Map the trigger time sequence to a set of pulse windows aligned with the time axis of the input AC signal, and write a window identifier associated with the zero-crossing reference time for each pulse window.

[0033] In this embodiment, based on the trigger time sequence generated in step S104, the distribution relationship of the trigger moments on the time axis is analyzed, and each trigger moment is mapped to the time axis position corresponding to the input AC signal, thereby dividing the time axis into several independent pulse windows. Specifically, with each trigger moment as the center or boundary, a corresponding time interval is determined on the time axis of the AC signal according to a preset time span rule, so that each time interval forms a pulse window. Subsequently, the zero-crossing reference moment determined in step S103 is used as a time anchor point and associated with the relative position relationship of each pulse window on the time axis, and a window identifier that reflects its correspondence with the zero-crossing reference moment is written for each pulse window. It should be noted that through the above mapping and identification process, the set of pulse windows is kept consistent with the input AC signal in the time dimension.

[0034] Step S1052: Perform window validity verification on the pulse window set, and remove windows that overlap in time or conflict in sequence with adjacent windows according to preset order constraint rules to obtain a pulse window set that has been constrained.

[0035] In this embodiment, the arrangement of each pulse window on the time axis is verified one by one for the pulse window set formed in step S1051 to confirm whether it meets the preset order constraint rules. Specifically, each pulse window is sorted according to its start time, and the time intervals of adjacent pulse windows are compared to identify window combinations with overlapping times or abnormal start and end sequences. For detected conflicts, the relevant pulse windows are eliminated or adjusted according to the order constraint rules, retaining only pulse windows that have a clear chronological order on the time axis and do not overlap. It should be noted that this legality verification process ensures that the retained pulse window set maintains consistency in time structure, avoiding temporal disorder caused by window conflicts in subsequent processing.

[0036] Step S1053: Using the constrained pulse window set as the boundary, the input AC signal is sampled and assigned values ​​in segments to form a pulse segment sequence corresponding to each pulse window, and the time interval outside the pulse window is set to a preset potential state.

[0037] In this embodiment, the constrained pulse window set obtained in step S1052 is used as the time boundary to segment the input AC signal on the time axis. Specifically, based on the start and end positions of each pulse window on the time axis, sampling is performed on the input AC signal within the corresponding time interval, and the sampling results are collected in chronological order to form a pulse segment sequence corresponding to each pulse window. For time intervals outside any pulse window, they do not participate in the formation of pulse segments, but are uniformly assigned values ​​according to preset potential rules, so that the time interval and the pulse window interval remain distinct in time structure. It should be noted that by using pulse windows as segmentation boundaries to separate the sampling and assignment processing of the AC signal, the pulse segment sequence can be made to strictly correspond to the pulse window set in the time dimension.

[0038] Step S1054: The pulse segment sequence is spliced ​​in time according to the window identifier, and a first pulse AC signal that maintains a time-series correspondence with the zero-crossing reference time is output.

[0039] In this embodiment, based on the pulse segment sequence and corresponding window identifiers formed in step S1053, the arrangement relationship of each pulse segment on the time axis is reconstructed. Specifically, according to the timing correlation information between the pulse segment and the zero-crossing reference time recorded in the window identifier, the pulse segment sequence is sorted and spliced ​​to ensure that the position of each pulse segment on the time axis is consistent with its corresponding pulse window. Subsequently, the sorted and spliced ​​pulse segments are combined with the preset potential segments outside the pulse windows in chronological order to form a continuous time sequence signal. It should be noted that by using the window identifier as the basis for timing splicing, the generated pulse AC signal can maintain a clear correspondence with the zero-crossing reference time in time structure, thereby obtaining the first pulse AC signal that meets the timing requirements of subsequent boost processing, i.e., the high-frequency pulse AC excitation signal.

[0040] like Figure 2 As shown, the input AC signal is a continuously changing periodic AC waveform. Multiple time points of the input AC signal are collected within a continuous sampling period along its time axis, forming a sampling data sequence reflecting the periodic changes of the AC signal. The sampling points are uniformly or quasi-uniformly distributed on the time axis to characterize the changes in the AC waveform within one cycle. Based on the sampling data sequence, the voltage sign changes of adjacent sampling points are analyzed. Multiple zero-crossing candidate segments are identified within the time neighborhood where the AC signal transitions from positive to negative polarity or vice versa, as shown by the shaded areas in the figure. Within each sampling period, the zero-crossing candidate segments are screened for time consistency to determine a unique zero-crossing reference time. In the figure, T1, T2, and T3 represent the zero-crossing reference times determined in different sampling periods, respectively.

[0041] After determining the zero-crossing reference time, a continuous phase evolution relationship is constructed using the zero-crossing reference time as a time base. According to the preset phase update rule, a trigger reference time is generated when the phase advances to a specified position, as shown in T1′, T2′, and T3′ in the figure. The trigger reference time maintains a fixed temporal correlation with the corresponding zero-crossing reference time on the time axis, which is used to limit the time position of subsequent pulsed processing.

[0042] Furthermore, based on the trigger reference time, pulsed processing is performed within the corresponding time interval of the input AC signal to convert the continuous AC waveform into discrete pulse segments. These pulse segments are then spliced ​​together according to the order of the trigger reference times on the time axis, thereby forming a first pulsed AC signal that maintains a timing correspondence with the zero-crossing reference time. Figure 2 As shown at the bottom.

[0043] Step S2: Input the first pulsed AC signal into a preset segmented capacitor multiplier network to obtain a second pulsating voltage signal. The segmented capacitor multiplier network is composed of multiple cascaded boost units, with a resonant branch set at the middle node to generate capacitor-inductor resonance within a preset frequency range, i.e., a high-amplitude pulsating high-voltage signal.

[0044] like Figure 3 As shown, the specific steps of step S2 are as follows: Step S201: Based on the first pulse AC signal, set the charging time slot sequence and transfer time slot sequence corresponding to the first pulse AC signal for multiple boost units, and assign a time slot identifier to each boost unit.

[0045] In this embodiment, the first pulse AC signal generated in step S1054 is used as a timing reference. Its pulse distribution on the time axis is analyzed, and corresponding charging time slot sequences and transfer time slot sequences are constructed for multiple boost units accordingly. Specifically, the pulse and non-pulse segments of the first pulse AC signal are distinguished on the time axis, and different time segments are mapped to time slot types suitable for charge writing or charge transfer according to preset timing allocation rules. Subsequently, based on the arrangement order of the boost units in the segmented capacitor multiplier network, the charging time slot sequence and transfer time slot sequence are allocated to each boost unit, and a corresponding time slot identifier is written for each boost unit to indicate the timing actions that the boost unit should perform in different time segments. It should be noted that by directly converting the timing characteristics of the first pulse AC signal into the time slot arrangement of the boost units, the charge processing processes of each boost unit form a coordinated relationship in the time dimension.

[0046] Step S202: Control the segmented capacitor multiplier network to enter a phased charging state according to the time slot identifier, so that each boost unit completes charge writing within the corresponding charging time slot and generates a charge state sequence.

[0047] In this embodiment, the working state of the segmented capacitor multiplier network in the time dimension is segmented and controlled according to the time slot identifiers assigned to each boost unit in step S201. Specifically, when a charging time slot arrives, the relevant boost unit enters the charge writing stage according to the corresponding time slot identifier, while boost units not identified as having charging time slots remain in a non-writing state. As the charging time slots advance along the time axis, each boost unit completes charge writing sequentially and records its charge retention status at the end of its respective charging time slot. Subsequently, the charge writing results of each boost unit in the corresponding charging time slot are summarized in chronological order to form a charge state sequence reflecting the segmented capacitor multiplier network during the staged charging process. It should be noted that by using the time slot identifier as the control basis, the charge writing process of each boost unit is distinguished from each other in time.

[0048] Step S203: Based on the charge state sequence, perform cross-cell charge transfer scheduling, and perform charge transfer and potential superposition processing on adjacent boost cells within the transfer time slot to form a node potential sequence, and write the node potential sequence into the node time stamp queue.

[0049] The specific steps of step S203 are as follows: Step S2031: Perform cell-level analysis on the charge state sequence, generate a migration candidate set of adjacent boost cell pairs according to the preset migration rules, and record the corresponding source cell, target cell and transfer time slot for each migration candidate.

[0050] In this embodiment, the charge state sequence formed in step S202 is analyzed at the unit level according to the arrangement of the boost units in the segmented capacitor multiplier network. Specifically, the charge state sequence is divided into state entries corresponding to each boost unit, and the state entries are paired and analyzed in combination with the order relationship between adjacent boost units. Subsequently, according to the preset migration rules, adjacent boost unit pairs that meet the migration conditions are marked, and a corresponding migration candidate entry is generated for each marked adjacent boost unit pair. For each migration candidate entry, its corresponding source unit, target unit, and transfer time slot for performing the migration operation are recorded, thereby forming a migration candidate set containing multiple sets of migration candidate information. It should be noted that the unit-level analysis of the charge state sequence and the generation of migration candidates are performed.

[0051] Step S2032: Perform conflict resolution processing on the migration candidate set. According to the preset mutual exclusion constraints, the migration candidates that share the same boost unit or the same transfer time slot are screened out and rearranged to obtain the migration execution sequence.

[0052] In this embodiment, the migration candidate set formed in step S2031 is analyzed from two dimensions: time and cell occupancy, to eliminate potential execution conflicts. Specifically, each candidate in the migration candidate set is first traversed to identify whether multiple candidates point to the same boost cell as the source or target cell, and whether multiple candidates are assigned to the same transfer time slot. For the detected conflicting candidates, they are processed according to a preset mutual exclusion constraint rule, which limits the same boost cell to only one charge transfer operation within any transfer time slot. Subsequently, migration candidates that do not meet the mutual exclusion constraint are screened out or their execution order is adjusted so that the remaining migration candidates do not conflict in terms of cell occupancy and time arrangement, and are sorted according to their time order to form a sequentially executable migration execution sequence. It should be noted that through this conflict resolution and rearrangement process, the charge transfer operation has a clear execution order on the time axis.

[0053] Step S2033: Within each transfer time slot corresponding to the migration execution sequence, charge migration and potential superposition operations are sequentially performed on the source unit and the target unit, and the instantaneous potential of the adjacent boost unit connection node is collected after each operation to form a node potential sequence arranged in time.

[0054] In this embodiment, the migration execution sequence obtained in step S2032 serves as a time guide, and charge migration and potential superposition operations are sequentially completed within each transfer time slot. Specifically, when a transfer time slot arrives, charge transfer is performed between the corresponding boost cells according to the source and target cell information recorded in the migration execution sequence, causing the charge already written in the source cell to migrate to the target cell, and causing the potential relationship between adjacent boost cells to superimpose during the migration process. After each charge migration and potential superposition operation is completed, the potential of the connection node between the source and target cells is acquired, and the acquisition result is associated with and stored in the current transfer time slot. As the migration execution sequence advances on the time axis, the node potential acquisition results in each transfer time slot are accumulated sequentially, thereby forming a node potential sequence arranged in chronological order. It should be noted that by acquiring the node potential after each migration operation, the node potential sequence can reflect the evolution of the node potential during the charge migration process.

[0055] It should be noted that charge migration is used to define the execution order and timing relationship of charge transfer between adjacent boost units. Its basic principles include: charge migration is only performed within the corresponding transfer time slot; within any transfer time slot, the same boost unit participates in at most one charge migration operation; the direction of charge migration follows the arrangement order of the boost units in the cascaded structure; after each charge migration is completed, a node potential acquisition operation is performed to update the node potential sequence.

[0056] Step S2034: Attach time tags corresponding to the transfer time slots to the node potential sequence and write them into the node time tag queue in time tag order.

[0057] In this embodiment, for the node potential sequence formed in step S2033, time correlation processing is performed on each node potential data and its corresponding transfer time slot. Specifically, each potential data in the node potential sequence is matched with the transfer time slot that generated the potential data, and a corresponding time tag is attached to it according to the position of the transfer time slot in the overall time sequence. Subsequently, the node potential data after attaching the tags are sorted according to the order of the time tags and written into the node time tag queue in sequence, so that the node time tag queue forms a continuous arrangement relationship in the time dimension. It should be noted that by attaching time tags to the node potential sequence and writing it into the node time tag queue in sequence, subsequent steps can have a clear time index when calling the node potential data, thereby realizing the time-series management and referencing of the node potential evolution process.

[0058] Step S204: Select the resonant intervention period corresponding to the node time stamp queue at the intermediate node of the segmented capacitor multiplier network, and introduce a resonant branch to participate in charge exchange during the resonant intervention period to generate a resonant participation identifier sequence.

[0059] The specific steps of step S204 are as follows: Step S2041: Divide the node time stamp queue into time slices, extract the node potential change segments corresponding to each time stamp, and convert each node potential change segment into a set of node change events.

[0060] In this embodiment, the node time-stamped queue formed in step S2034 is used as the time-series input, and the queue is segmented in the time dimension. Specifically, based on the interval relationship between adjacent time stamps in the node time-stamped queue, the queue is divided into several consecutive time slices, and the corresponding node potential data is extracted within each time slice. Subsequently, the node potential data within the same time slice is compared before and after to identify the changes in node potential within that time slice, and these changes are summarized into a node potential change segment. For each node potential change segment, based on its position and change characteristics on the time axis, it is converted into a node change event with clear time boundaries and change attributes, thereby forming a set of node change events. It should be noted that by dividing the node time-stamped queue into time slices and extracting node potential change segments, the process of node potential change is expressed in the form of events.

[0061] Step S2042: Filter the set of node change events, determine the target events that meet the triggering conditions according to the preset resonance triggering rules, and assign a corresponding resonance intervention time period identifier to each target event to form a set of resonance intervention time periods.

[0062] In this embodiment, based on the node change event set formed in step S2041, each node change event is filtered in terms of both time position and change characteristics. Specifically, each event in the node change event set is arranged according to its time tag order, and the node potential change characteristics reflected by each change segment are compared to see if they conform to a preset resonance triggering rule. The resonance triggering rule is used to limit the node change event to a candidate event for resonance intervention only if it meets specific conditions in terms of time continuity, change direction, or change amplitude range. For node change events that meet the triggering conditions, a corresponding resonance intervention period identifier is assigned to them according to their time position in the node time tag queue, and the resonance intervention period identifier is associated with the corresponding node change event. Node change events that do not meet the triggering conditions do not participate in subsequent processing. Through the above filtering and identifier process, a resonance intervention period set is formed that only contains events that meet the resonance triggering conditions.

[0063] It should be noted that the resonance triggering rule includes at least one or more of the following judgment conditions: the node potential shows a continuous changing trend within adjacent time slices; the direction of the node potential change is consistent with the expected charge migration direction; the magnitude of the node potential change is within the allowable change range; the node potential change event is located within a predetermined time window after the charge migration of the adjacent boost unit is completed; when the node change event meets the above conditions, it is determined to be a target event that allows the resonance branch to intervene, and a corresponding resonance intervention period is assigned to it.

[0064] Step S2043: Within each resonance intervention period defined by the set of resonance intervention periods, perform resonance branch intervention operation, record the state of the charge exchange process of the intermediate node within the resonance intervention period, and generate resonance participation record corresponding to each resonance intervention period identifier.

[0065] In this embodiment, the set of resonant intervention time periods formed in step S2042 is used as a time constraint to perform time-sequential control on the intervention process of the resonant branch in the segmented capacitor multiplier network. Specifically, when the time axis advances to the time interval defined by any resonant intervention time period identifier, the corresponding resonant branch intervention operation is executed, so that the intermediate node participates in the resonant-related charge exchange process within that time interval. During the resonant intervention time period, the charge exchange state of the intermediate node is continuously recorded, and the recorded state information is used to characterize the temporal relationship between the node potential and charge change during the resonant intervention. Subsequently, the recorded state information is associated and stored with the corresponding resonant intervention time period identifier to form a resonant participation record corresponding one-to-one with each resonant intervention time period.

[0066] Step S2044: Encode the resonance participation record and output the resonance participation identifier sequence in chronological order. The resonance participation identifier sequence is used to characterize the participation status of the resonance branch in each resonance intervention period.

[0067] In this embodiment, the time period identifiers and state information contained in each record item of the resonance participation record formed in step S2043 are organized and abstracted. Specifically, the state content reflecting the intervention of the resonant branch in each resonance participation record is extracted and converted into a distinguishable identifier form according to a preset encoding rule. Subsequently, the encoded identifiers are sorted and arranged according to the chronological order of each resonance intervention time period on the time axis to form a continuous resonance participation identifier sequence. It should be noted that by encoding the resonance participation record and outputting the identifier sequence in chronological order, the participation state of the resonant branch in different intervention time periods can be expressed in a serialized manner, thereby providing a clear state input for subsequent time-series updates and signal processing based on the identifier sequence.

[0068] Step S205: Update the node potential sequence in time according to the resonant participation identifier sequence within a preset frequency range, and output a second pulsating voltage signal corresponding to the node potential sequence.

[0069] The specific steps of step S205 are as follows: Step S2051: Perform time alignment processing on the resonance participation identifier sequence, and map the resonance participation identifier sequence to the time tag of the node potential sequence to obtain the resonance mapping sequence.

[0070] In this embodiment, a timing alignment process is performed on the temporal relationship between the resonance participation identifier sequence formed in step S2044 and the node potential sequence formed in step S2034. Specifically, firstly, the time information of the resonance intervention period corresponding to each identifier in the resonance participation identifier sequence is extracted, and this time information is compared with the existing time tags in the node potential sequence. Subsequently, based on the correspondence of the time tags on the time axis, each resonance participation identifier is mapped to the node potential data position that coincides with or is closest to its time interval, and the association between the two is established. Through the above mapping process, the resonance participation identifier sequence and the node potential sequence form a one-to-one correspondence structure in the time dimension, thereby obtaining a resonance mapping sequence containing node potential changes and resonance participation state information.

[0071] Step S2052: Based on the resonance mapping sequence, the node potential sequence is divided into resonance update segments and non-resonance update segments, and update rule sets corresponding to each segment are generated respectively.

[0072] In this embodiment, the resonance mapping sequence obtained in step S2051 is used as the discrimination criterion to classify the node potential sequence in the time dimension. Specifically, by analyzing the resonance participation state corresponding to each time tag in the resonance mapping sequence, the time intervals in the node potential sequence corresponding to the resonance intervention period are identified, and these time intervals are divided into resonance update segments. The remaining time intervals not associated with the resonance participation state are divided into non-resonance update segments. Subsequently, for different types of update segments, corresponding update rule sets are generated. The update rules for resonance update segments are used to limit the processing method of node potentials within this type of time interval, while the update rules for non-resonance update segments are used to limit their processing method in other time intervals.

[0073] Step S2053: Within the preset frequency range, each resonant update segment and each non-resonant update segment are sequentially updated according to the update rule set to obtain the updated node potential sequence.

[0074] In this embodiment, the update rule set generated in step S2052 is used as the processing basis to perform sequential updates on the node potential sequence within a preset frequency range. Specifically, firstly, based on the chronological relationship of each update segment in the node potential sequence on the time axis, the resonant update segments and non-resonant update segments are sorted, and the update rule set corresponding to the current segment type is called sequentially during the time progression. For node potential data within a resonant update segment, the node potential state is reconstructed according to the rules corresponding to that type of segment; for node potential data within a non-resonant update segment, it is processed according to another set of rules. As the time axis continuously advances within the preset frequency range, each segment completes the update operation in sequence, and finally, the updated segments are reassembled to form an updated node potential sequence continuously arranged in the time dimension.

[0075] It should be noted that the preset frequency range is determined based on the fundamental frequency of the input AC signal or its integer multiples; it should meet the following requirements: it can fully cover one or more AC sampling cycles; it can ensure that each boost unit completes at least one charging or charge transfer operation within one cycle; and it can be adjusted according to the actual application scenario, load conditions, and device withstand voltage capability.

[0076] Step S2054: Perform timing shaping on the updated node potential sequence and output the second pulsating voltage signal corresponding to the updated node potential sequence.

[0077] The specific steps of step S2054 are as follows: Step S20541: Divide the updated node potential sequence into segments according to time labels, extract potential change segments between adjacent time labels to form a segment set, and write a segment identifier for each segment.

[0078] In this embodiment, the updated node potential sequence obtained in step S2053 is re-divided and re-identified along the time dimension. Specifically, based on the time tags attached to each potential data in the node potential sequence, the sequence is segmented in chronological order, and the range of node potential changes between adjacent time tags is extracted as independent potential change segments. For each potential change segment, a unique segment identifier is written according to its positional order on the time axis and its corresponding time span, which is used to characterize the relative position of the segment in the overall node potential evolution process.

[0079] Step S20542: Perform boundary consistency processing on the segment set, fill in the segments with gaps according to the preset boundary rules, and trim the segments with overlaps to obtain the segment set after boundary processing.

[0080] In this embodiment, for the segment set formed in step S20541, the continuity of each potential change segment on the time boundary is consistent. Specifically, the start and end times of adjacent segments in the segment set are compared to identify segments with time gaps or overlaps. For segments with time gaps, according to preset boundary rules, supplementary segments are inserted between adjacent segments or the time boundaries of adjacent segments are extended to eliminate gaps on the time axis. For segments with overlapping time boundaries, the overlapping parts are trimmed according to the order of the segment identifiers to ensure that the segments do not overlap on the time axis. Through the above supplementary and trimming operations, the segment set forms a continuous and non-overlapping arrangement in the time structure, thereby obtaining a segment set with boundary processing.

[0081] Step S20543: Perform sequence arrangement processing on the set of segments after boundary processing, rearrange the segment identifiers according to the preset time arrangement rules, and splice the rearranged segments in the corresponding order to form the time-shaped node potential sequence.

[0082] In this embodiment, based on the boundary-processed segment set obtained in step S20542, the arrangement of each segment on the time axis is reorganized. Specifically, according to a preset time sequence arrangement rule, the temporal sequence relationship corresponding to each segment identifier in the segment set is parsed, and in the case of inconsistent order or redundant arrangement, the segment identifiers are rearranged to conform to a unified time progression logic. Subsequently, according to the rearranged segment identifier order, the corresponding potential change segments are sequentially spliced ​​together, so that adjacent segments are connected end-to-end on the time axis to form a continuous node potential change process.

[0083] Step S20544: Based on the timing-shaped node potential sequence, output a second pulsating voltage signal corresponding to the timing-shaped node potential sequence.

[0084] In this embodiment, the time-shaped node potential sequence formed in step S20543 is used as the basis for signal generation, and the continuous change relationship of this sequence on the time axis is mapped. Specifically, the potential changes corresponding to each time segment in the node potential sequence are output and configured according to their time order, so that the arrangement relationship of the node potentials in the time dimension is completely preserved and converted into the corresponding voltage change form. Subsequently, based on the time structure of the node potential sequence, the potential changes of adjacent time segments are continuously connected to form a pulsating voltage expression with a clear time order. It should be noted that by using the time-shaped node potential sequence as the sole reference source for output mapping, the generated pulsating voltage signal is kept consistent with the node potential change process in time structure, thereby obtaining a second pulsating voltage signal corresponding to the time-shaped node potential sequence.

[0085] like Figure 3 As shown, a first pulse AC signal is introduced as an input signal into a segmented capacitor multiplier network composed of multiple cascaded boost units. The segmented capacitor multiplier network includes multiple boost units arranged sequentially, corresponding to S1, S2, S3, and S4 as indicated in the figure. Each boost unit participates in the energy conversion process on the time axis according to a unified timing rule. Based on the pulse distribution characteristics of the first pulse AC signal on the time axis, a corresponding charging time slot sequence is set for each boost unit, as shown in the charging time slot sequence in the figure. When the charging time slot arrives, the corresponding boost unit is controlled to enter the charging state, enabling the boost unit to complete the charge writing operation. As the charging time slot advances on the time axis, each boost unit completes the charging process sequentially and forms a corresponding charge holding state after each charging time slot ends, thereby generating the charge state sequence Q1, Q2, Q3, and Q4 shown below the figure.

[0086] After completing the phased charging, based on the charge state sequence, cross-unit charge migration scheduling is performed on adjacent boost units within a preset transfer time slot, as shown in P1, P2, P3, and P4 in the figure. Within the corresponding transfer time slot, the charge stored in the previous boost unit is transferred to the subsequent boost unit, and a potential superposition relationship is formed during the transfer process. After each charge migration is completed, the potential of the connection node between adjacent boost units is collected, so that the potential of each node forms a node potential sequence in time order.

[0087] Furthermore, at the intermediate node of the segmented capacitor multiplier network, a resonant intervention period is selected according to the node potential sequence, and a resonant branch is introduced to participate in the charge exchange process during the resonant intervention period, as shown in the resonant branch structure on the right side of the figure. During the resonant intervention period, the charge exchange state of the intermediate node is recorded, and the corresponding participation status is encoded to form a resonant participation identifier sequence.

[0088] Subsequently, within a preset frequency range, the node potential sequence is updated in a time sequence according to the resonant participation identifier sequence, so that the node potential changes in an orderly manner over time. The updated node potential sequence is continuously output on the time axis, and finally forms the second pulsating voltage signal as shown on the right side of the figure. This second pulsating voltage signal is used as the input signal for the subsequent DC collection and feedback regulation steps.

[0089] Step S3: The second pulsating voltage signal is DC collected, and the parameters of the first pulse AC signal are adjusted according to the voltage status feedback of each boost unit to generate a stable kilovolt-level DC voltage.

[0090] The specific steps of step S3 are as follows: Step S301: Perform unidirectional processing on the second pulsating voltage signal according to the preset rectification timing, and write the unidirectional signal into the DC collection sequence.

[0091] In this embodiment, the second pulsating voltage signal output in step S20544 is used as the processing object. The pulsating voltage signal is unidirectionally processed in the time dimension according to a preset rectification timing sequence. Specifically, the voltage polarity of the second pulsating voltage signal in each time segment is determined, and according to the time constraint of the rectification timing sequence, only voltage segments that conform to the preset direction constraint are retained, while segments that do not conform to the direction constraint are masked or equivalently replaced. Subsequently, each voltage segment after unidirectional processing is written into the DC collection sequence in chronological order, so that the sequence forms a continuously arranged unidirectional voltage data on the time axis. It should be noted that by performing unidirectional processing on the second pulsating voltage signal at the time sequence level, rather than performing instantaneous rectification judgment, the DC collection sequence maintains consistency in its time structure.

[0092] Step S302: Perform hierarchical sampling on the DC collection sequence to obtain a voltage state set corresponding to each boost unit, and write the voltage state set into the state time stamp queue.

[0093] In this embodiment, for the DC-DC aggregation sequence formed in step S301, hierarchical sampling processing is performed in two dimensions: time and unit level. Specifically, based on the correspondence of each boost unit in the segmented capacitor multiplier network, the DC-DC aggregation sequence is divided into several sampling levels, and voltage sampling is performed at the time position associated with the boost unit within each level. Subsequently, the sampling results are categorized by boost unit to form a voltage state set that reflects the current voltage state of each boost unit. For the formed voltage state set, a time tag corresponding to the sampling time is attached to it, and it is written into the state time stamp queue in chronological order.

[0094] Step S303: Construct a parameter update sequence based on the state time stamp queue. The parameter update sequence includes update instructions for the pulse width parameter, pulse interval parameter, and pulse cluster length parameter of the first pulse AC signal.

[0095] In this embodiment, the state time-stamped queue formed in step S302 is used as input to perform correlation analysis on the voltage states of each boost unit under different time tags. Specifically, the voltage state set in the state time-stamped queue is read in time tag order, and the voltage states corresponding to adjacent time tags are compared to identify the changing trend of the voltage state in the time dimension. Subsequently, according to the preset parameter correlation rules, the changing trend is mapped to the adjustment requirements of the first pulse AC signal parameters, and update instructions for the pulse width parameter, pulse interval parameter, and pulse cluster length parameter are generated respectively. The update instructions are arranged in chronological order to form a parameter update sequence.

[0096] Step S304: Iteratively adjust the parameters of the first pulse AC signal according to the parameter update sequence, and output a kilovolt-level DC voltage corresponding to the DC collection sequence.

[0097] In this embodiment, the parameter update sequence generated in step S303 is used as the adjustment basis to iteratively update the parameters of the first pulse AC signal. Specifically, according to the time sequence indicated in the parameter update sequence, the pulse width parameter, pulse interval parameter, and pulse cluster length parameter are adjusted sequentially, causing the time structure of the first pulse AC signal to change accordingly in subsequent cycles. As the parameter update sequence is executed item by item, the adjusted first pulse AC signal participates again in the aforementioned boosting, collection, and sampling process, continuously updating the DC collection sequence in the time dimension. It should be noted that by cyclically associating the parameter update sequence with the generation process of the first pulse AC signal, the parameter adjustment of the pulse AC signal and the change of the DC collection sequence form an iterative relationship, thereby outputting a kilovolt-level DC voltage corresponding to the DC collection sequence during the iteration process.

[0098] Example 2 Please see Figure 2 Another embodiment of the present invention provides: an AC-to-DC boosting power conversion device for kilovolts, comprising: a first signal generation module, a second signal generation module, and a kilovolt voltage output module; The first signal generation module is used to acquire the input AC signal and generate a first pulse AC signal that is phase-locked with the zero-crossing point of the AC signal based on the acquired input AC signal. The second signal generation module is used to input the first pulse AC signal into a preset segmented capacitor voltage multiplier network to obtain a second pulsating voltage signal. The segmented capacitor voltage multiplier network is composed of multiple cascaded boost units, and a resonant branch is set in the middle node to generate capacitor-inductor resonance within a preset frequency range. The kilovolt voltage output module is used to collect the second pulsating voltage signal into DC and adjust the parameters of the first pulse AC signal according to the voltage status feedback of each boost unit to generate a stable kilovolt-level DC voltage.

[0099] In addition, the parts of the technical solutions provided in the embodiments of this application that are consistent with the implementation principles of the corresponding technical solutions in the prior art have not been described in detail, so as to avoid excessive elaboration.

[0100] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for AC-to-DC voltage boosting to kilovolts at the local end, characterized in that, The method is applied to DC high-voltage aging power supply conversion scenarios, enabling the conversion of low-voltage AC power to kilovolt-level DC power without using a power frequency core transformer. Specifically, it includes: The system acquires the input AC signal and generates a first pulse AC signal that is phase-locked with the zero-crossing point of the AC signal based on the acquired input AC signal. The first pulsed AC signal is input into a preset segmented capacitor multiplier network to obtain a second pulsating voltage signal. The segmented capacitor multiplier network is composed of multiple cascaded boost units, with a resonant branch set in the middle node to generate capacitor-inductor resonance within a preset frequency range. The second pulsating voltage signal is DC collected, and the parameters of the first pulse AC signal are adjusted according to the voltage status feedback of each boost unit to generate a stable kilovolt-level DC voltage.

2. The AC-to-DC boosting to kilovolt local-end energy conversion method as described in claim 1, characterized in that, The process of acquiring the input AC signal and generating a first pulse AC signal that is phase-locked with the zero-crossing point of the AC signal based on the acquired input AC signal includes: Within a continuous sampling period, the input AC signal is acquired at multiple time points to obtain a sampling data sequence that reflects the periodic changes of the AC signal. Perform sign change analysis on the sampled data sequence to identify time segments in the sampled data where the voltage polarity changes, and form a set of zero-crossing candidate segments; The zero-crossing candidate segment set is subjected to time consistency screening to determine the zero-crossing reference time corresponding to the current sampling period, and a time identifier corresponding to the zero-crossing reference time is established; A continuous phase evolution relationship is constructed based on the time identifier, and a trigger time sequence is generated according to a preset phase update rule; Based on the trigger time sequence, pulse processing is performed within the corresponding time interval of the input AC signal to generate a first pulse AC signal that maintains a timing correspondence with the zero-crossing reference time.

3. The AC-to-DC boosting to kilovolt local-end energy conversion method as described in claim 2, characterized in that, Based on the trigger time sequence, pulse processing is performed within the corresponding time interval of the input AC signal to generate a first pulse AC signal that maintains a timing correspondence with the zero-crossing reference time, including: The trigger time sequence is mapped to a set of pulse windows aligned with the time axis of the input AC signal; The set of pulse windows is subjected to window validity verification, and windows that overlap in time or conflict in sequence with adjacent windows are removed according to preset order constraint rules to obtain a set of pulse windows that have been constrained. Using the constrained pulse window set as the boundary, the input AC signal is segmented and sampled and segmented to form a pulse segment sequence corresponding to each pulse window. The pulse segment sequence is time-sequentially spliced ​​according to the window identifier, and a first pulse AC signal that maintains a time-sequential correspondence with the zero-crossing reference time is output.

4. The AC-to-DC boosting to kilovolt local-end energy conversion method as described in claim 3, characterized in that, The first pulsed AC signal is input into a preset segmented capacitor multiplier network to obtain a second pulsating voltage signal, including: Based on the first pulse AC signal, a charging time slot sequence and a transfer time slot sequence corresponding to the first pulse AC signal are set for multiple boost units, and a time slot identifier is assigned to each boost unit; According to the time slot identifier, the segmented capacitor multiplier network is controlled to enter a phased charging state, so that each boost unit completes charge writing within the corresponding charging time slot and generates a charge state sequence. Based on the charge state sequence, cross-cell charge transfer scheduling is performed. During the transfer time slot, adjacent boost cells are subjected to charge transfer and potential superposition processing to form a node potential sequence, and the node potential sequence is written into the node time stamp queue. In the segmented capacitor multiplier network, a resonant intervention period corresponding to the node time stamp queue is selected at the intermediate node, and a resonant branch is introduced to participate in charge exchange during the resonant intervention period to generate a resonant participation identifier sequence. Within a preset frequency range, the node potential sequence is updated in time according to the resonant participation identifier sequence, and a second pulsating voltage signal corresponding to the node potential sequence is output.

5. The AC-to-DC boosting to kilovolt local-end energy conversion method as described in claim 4, characterized in that, Based on the charge state sequence, cross-cell charge transfer scheduling is performed. During the transfer time slot, adjacent boost cells undergo charge transfer and potential superposition processing to form a node potential sequence, which is then written into the node time stamp queue. This includes: The charge state sequence is analyzed at the cell level, and a migration candidate set of adjacent boost cell pairs is generated according to a preset migration rule. The source cell, target cell and transfer time slot are recorded for each migration candidate. The migration candidate set is subjected to conflict resolution processing. According to the preset mutual exclusion constraints, migration candidates that share the same boost unit or share the same transfer time slot are screened out and rearranged to obtain the migration execution sequence. Within each transfer time slot corresponding to the migration execution sequence, charge migration and potential superposition operations are sequentially performed on the source unit and the target unit, and the instantaneous potential of the adjacent boost unit connection node is collected after each operation to form a node potential sequence arranged in time. The node potential sequence is appended with a time tag corresponding to the transfer time slot, and written into the node time tag queue in the order of the time tags.

6. The AC-to-DC boosting to kilovolt local-end energy conversion method as described in claim 5, characterized in that, In the segmented capacitor multiplier network, a resonant intervention period corresponding to the node time-stamped queue is selected at the intermediate node, and a resonant branch is introduced to participate in charge exchange within the resonant intervention period to generate a resonant participation identifier sequence, including: The node time stamp queue is divided into time slices, the node potential change segments corresponding to each time stamp are extracted, and each node potential change segment is converted into a set of node change events. The set of node change events is filtered to determine target events that meet the triggering conditions according to preset resonance triggering rules, and each target event is assigned a corresponding resonance intervention time period identifier to form a set of resonance intervention time periods. Within each resonance intervention period defined by the set of resonance intervention periods, resonance branch intervention operation is performed, and the state of the charge exchange process of the intermediate node during the resonance intervention period is recorded to generate resonance participation record corresponding to each resonance intervention period identifier. The resonance participation record is encoded and a resonance participation identifier sequence is output in chronological order. The resonance participation identifier sequence is used to characterize the participation status of the resonance branch during each resonance intervention period.

7. The AC-to-DC boosting to kilovolt central office energy conversion method as described in claim 6, characterized in that, The step of updating the node potential sequence in a time series according to the resonant participation identifier sequence within a preset frequency range and outputting a second pulsating voltage signal corresponding to the node potential sequence includes: The resonance participation identifier sequence is time-aligned and mapped to the time label of the node potential sequence to obtain the resonance mapping sequence. Based on the resonant mapping sequence, the node potential sequence is divided into resonant update segments and non-resonant update segments, and update rule sets corresponding to each segment are generated respectively. Within the preset frequency range, each resonant update segment and each non-resonant update segment are sequentially updated according to the set of update rules to obtain the updated node potential sequence. The updated node potential sequence is subjected to timing shaping, and a second pulsating voltage signal corresponding to the updated node potential sequence is output.

8. The AC-to-DC boosting to kilovolt local-end energy conversion method as described in claim 7, characterized in that, The updated node potential sequence is subjected to timing shaping, and a second pulsating voltage signal corresponding to the updated node potential sequence is output, including: The updated node potential sequence is segmented according to time labels, and potential change segments between adjacent time labels are extracted to form a segment set. The segment set is subjected to boundary consistency processing. Segments with gaps are filled and segments with overlaps are trimmed according to preset boundary rules to obtain a segment set with boundary processing. The set of segments after boundary processing is subjected to sequence arrangement processing. The segment identifiers are rearranged according to the preset time arrangement rules, and the rearranged segments are spliced ​​together in the corresponding order to form a time-shaped node potential sequence. Based on the time-shaped node potential sequence, a second pulsating voltage signal corresponding to the time-shaped node potential sequence is output.

9. The AC-to-DC boosting to kilovolt local-end energy conversion method as described in claim 8, characterized in that, The second pulsating voltage signal is DC collected, and the parameters of the first pulsed AC signal are adjusted according to the voltage status feedback of each boost unit to generate a stable kilovolt-level DC voltage, including: The second pulsating voltage signal is unidirectionally processed according to a preset rectification timing, and the unidirectional signal is written into the DC converging sequence. The DC collection sequence is sampled hierarchically to obtain a voltage state set corresponding to each boost unit, and the voltage state set is written into a state time stamp queue. A parameter update sequence is constructed based on the state time stamp queue. The parameter update sequence includes update instructions for the pulse width parameter, pulse interval parameter, and pulse cluster length parameter of the first pulse AC signal. The parameters of the first pulse AC signal are iteratively adjusted according to the parameter update sequence, and a kilovolt-level DC voltage corresponding to the DC collection sequence is output.

10. An AC-to-DC boosting power conversion device for local-end voltage conversion, used to implement the AC-to-DC boosting power conversion method for local-end voltage conversion as described in any one of claims 1-9, characterized in that, include: The system comprises a first signal generation module, a second signal generation module, and a kilovolt voltage output module. The first signal generation module is used to acquire the input AC signal and generate a first pulse AC signal that is phase-locked with the zero-crossing point of the AC signal based on the acquired input AC signal. The second signal generation module is used to input the first pulse AC signal into a preset segmented capacitor voltage multiplier network to obtain a second pulsating voltage signal. The segmented capacitor voltage multiplier network is composed of multiple cascaded boost units, and a resonant branch is set in the middle node to generate capacitor-inductor resonance within a preset frequency range. The kilovolt voltage output module is used to collect the second pulsating voltage signal into DC and adjust the parameters of the first pulse AC signal according to the voltage status feedback of each boost unit to generate a stable kilovolt-level DC voltage.

Images