A charger testing method and system

Abstract

The application relates to the field of electrical measurement and electronic test equipment, and discloses a charger test method and system. The method comprises the following steps: acquiring a parameter configuration set and a baseline condition, generating a probe and interface configuration structure under the conditions of a four-wire measurement site, protocol abstract card mapping and shunt and Hall channel combination; performing unified time reference binding and event label table initialization, locking a time base by a real-time clock of a microcontroller unit and broadcasting to channels and controllers, executing temperature rise constraint load arrangement and synchronous acquisition, and generating an acquisition preparation package; and executing electrical transient and thermal characteristic extraction and electrical-thermal decoupling fusion, completing consistency determination, threshold comparison and report generation. Through the establishment of a unified time base structure for electrical and thermal dual-domain synchronization, the application realizes multi-channel consistent acquisition and real-time self-diagnosis, improves the precision, traceability and automation level of the charger test process, and is suitable for comprehensive testing and evaluation of multiple types of electronic equipment.

Description

Technical Field

[0001] This invention relates to the field of electrical measurement and electronic testing equipment, and more particularly to a charger testing method and system. Background Technology

[0002] In the field of electrical parameter testing and thermal characteristic evaluation, existing charger testing solutions typically rely on multi-channel sampling devices, protocol adaptation units, and load control equipment to form a test link. This presents limitations such as separation of electrical and thermal channels, asynchronous time bases, and discontinuous judgment processes. Existing methods often employ separate acquisition devices and step-by-step control logic, which are prone to measurement drift and event alignment failures under temperature rise constraints and dynamic load changes, making it difficult to meet the stable and consistent testing requirements of chargers under multiple protocols and operating conditions. Regarding the joint processing of parameter configuration sets, probe and interface configuration structures, and unified time base binding, existing technologies generally suffer from insufficient coordination in the synchronous acquisition of electrical transient characteristics and thermal responses, event labeling, and threshold comparison. This makes it difficult to form a closed-loop link of acquisition, alignment, judgment, and updating under a unified time base, leading to problems such as cross-stage test data mismatch, delayed threshold updates, and interrupted configuration feedback. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a charger testing method, comprising: The parameter configuration set and baseline conditions are obtained, and the zero-disturbance segment acquisition and constant micro-carrier segment acquisition are performed. Under the conditions of synchronous execution of four-wire Kelvin site registration and protocol abstract card mapping, shunt and Hall channel combination and temperature-sensitive site arrangement, the probe and interface configuration structure is generated. The process includes binding a unified time base and initializing the event label table; locking the time base with the real-time clock (RTC) of the microcontroller unit and broadcasting it to each channel; executing and synchronizing the temperature rise constraint load orchestrator of the protocol abstraction card and the electronic load controller; performing probe health self-checks and disturbance labeling; and generating the acquisition preparation package structure. Perform electrical transient feature extraction and ripple and overshoot fallback processing, calculate temperature gradient, interval equivalent thermal resistance and thermal inertia duration under the same time base, and perform electrothermal decoupling fusion to generate evaluation input packet structure; Perform baseline-incentive-regression consistency determination and threshold comparison, anomaly three-part diagnosis and secondary verification script arrangement, uncertainty item budgeting and standardization report generation and configuration threshold update processing, and generate configuration and threshold update records.

[0004] Furthermore, the process of binding a unified time base includes: Unified time base binding refers to the process of establishing a unified time base among all electrical and thermal sampling channels, protocol abstraction card handshake process, and electronic load actions. The time base is locked by the real-time clock (RTC) of the microcontroller unit and broadcast to the channel scheduler, protocol abstraction card adaptation layer, and electronic load controller. After the broadcast is completed, the sampling board where the analog-to-digital converter (ADC) is located sends back a time base confirmation frame, which is recorded in the time base confirmation column of the event label table.

[0005] Furthermore, the process of initializing the event label table includes: The event label table is initialized by creating a structured table that grows in chronological order. Each table entry must contain at least the fields for time stamp, event type, event source, associated channel, handshake phase, load action, and exception flag. During initialization, time base lock, channel ready, protocol abstract card ready, and load controller ready are written.

[0006] Furthermore, the temperature rise constrained load orchestrator includes: the temperature rise constrained load orchestrator is a control unit that orchestrates load instructions based on dual constraints of electrical and thermal states. Its inputs include the abstract instruction set, channel mapping, recording strategy image and stable segment set provided by the protocol abstraction card, and its outputs are electronic load gear switching sequence and protocol handshake advancement instructions.

[0007] Furthermore, the synchronous acquisition process also includes: Synchronous acquisition involves parallel sampling of the voltage channel, current channel, and temperature-sensitive channel under the same time base while the load action and handshake phases are progressing. The sampling scheduling follows the channel priority relationship in the probe and interface configuration structure, realizing current shunting and Hall parallel sampling, voltage Kelvin dual-point parallel sampling, and thermal channel number sequential sampling.

[0008] Furthermore, the determination of temperature rise constraint is driven by the temperature gradient of the thermal channel and the temperature rise rate of adjacent sampling windows. When the temperature rise rate exceeds the gradient threshold in the recording strategy image, the load orchestrator writes a temperature rise over-limit warning event and reduces the load amplitude. When the temperature approaches the upper bound of the stable segment set, a thermal equilibrium dwell event is inserted.

[0009] Furthermore, the probe health self-test performs three types of inspections within the sampling window: open circuit judgment, short circuit judgment, and reading drift judgment. These are based on the low-power probe pulse response, port differential voltage, and continuous window trend offset, respectively. When an abnormality is detected, a probe self-test abnormality is written and the channel switching logic is driven.

[0010] Furthermore, the channel switching logic switches to the backup channel when a backup channel exists and writes the channel switching to take effect. When no backup channel exists, it sets a downgrade operation flag for the site and reduces its sampling priority. All results are synchronously written back to the recording strategy mirror.

[0011] Furthermore, the disturbance labeling process identifies and labels external and internal disturbances. External disturbances include power supply jitter, sudden changes in on-site wind speed, and ambient temperature steps. Internal disturbances include electronic load quantization steps, channel sampling jitter, and protocol abstraction card handshake retries. The corresponding disturbance labels are written into the event labeling table.

[0012] Furthermore, a charger testing system, applied to any of the above methods, includes: The probe and interface configuration unit is used to perform zero-disturbance segment acquisition and constant microcarrier segment acquisition, four-wire Kelvin site registration, protocol abstract card mapping, shunt and Hall channel combination, and environmental and internal temperature-sensitive site arrangement processing according to the parameter configuration set and baseline conditions, and generate the probe and interface configuration structure. The time base and event labeling unit is used to load the probe and interface configuration structure, complete the unified time base binding and event labeling table initialization in the test main control unit, and establish an event structure that grows in chronological order. The load orchestration and synchronous acquisition unit is used to perform load orchestration and synchronous acquisition processing of progressive excitation sequence according to the protocol abstract card instruction set and temperature rise constraint conditions, and record load actions and sampling events under a unified time base. The probe health and disturbance determination unit is used to perform probe health self-check and disturbance labeling processing during the acquisition process, and to identify and label the probe contact status, channel stability and external and internal disturbances; The feature extraction and fusion unit is used to perform electrical transient feature extraction, ripple and overshoot fallback processing, temperature gradient and equivalent thermal resistance and thermal inertia duration calculation, and electrothermal decoupling fusion processing on the sampled data of voltage channel, current channel and temperature-sensitive channel based on the acquisition preparation packet structure, and generate evaluation input packet structure. The consistency determination and reporting unit is used to call the evaluation input package structure, perform baseline-excitation-regression consistency determination and threshold comparison, anomaly three-part diagnosis and secondary verification script orchestration, uncertainty item budgeting and standardized report generation, and generate configuration and threshold update records; The threshold and parameter update unit is used to write the threshold domain and fixture review suggestions of the configuration and threshold update record into the parameter configuration set, providing loading input for the next batch of tests and realizing the closed loop of the test process.

[0013] The key innovations of this invention include: (1) The processing links of zero-disturbance segment acquisition and constant micro-carrier segment acquisition, registration at four-wire Kelvin site and mapping of protocol abstract card, shunting and Hall channel combination and temperature-sensitive site arrangement are organized in the same generation path to directly generate probe and interface configuration structure.

[0014] (2) The unified time base binding, event label table initialization, temperature rise constraint load orchestrator execution, and synchronous acquisition are carried out in parallel and coordinated, and probe health self-check and disturbance label processing are combined under the same time base to generate the acquisition preparation package structure by stage merging.

[0015] (3) Under the same time base, the extraction of electrical transient features and the processing of ripple and overshoot fallback are electrothermally decoupled and fused with temperature gradient, interval equivalent thermal resistance and thermal inertia duration calculation. The output evaluation input packet structure is then formed into a closed loop with the baseline-excitation-regression consistency judgment and threshold comparison, standardized report generation and configuration threshold update processing.

[0016] The following are its main beneficial effects: (1) The generation process of probe and interface configuration structure completes the physical measurement chain, protocol adaptation chain and thermal observation chain in the same transaction, reducing the risk of interface inconsistency and channel switching omission caused by step registration and multiple mappings in the existing scheme, and facilitating subsequent alignment and judgment on a unified object.

[0017] (2) Parallel collaboration oriented to the acquisition preparation package structure enables load actions, sampling frames and event labels to form a continuous index under a unified time base, which alleviates the data mismatch problem caused by time base asynchrony and event alignment failure in the existing scheme, and facilitates the stable acquisition of time series data that can be used for judgment under dynamic load and temperature rise constraint scenarios.

[0018] (3) To evaluate the linkage between the input packet structure and the configuration and threshold update records, the electrical transient characteristics and thermal entries are directly entered into the baseline-excitation-regression consistency judgment and threshold comparison after electrothermal decoupling and fusion, and the results are written back to the configuration threshold update process, which improves the configuration feedback interruption problem caused by the discontinuous judgment process and the lag in threshold update in the existing scheme. Attached Figure Description

[0019] Figure 1 A flowchart illustrating a charger testing method provided in an embodiment of this application; Figure 2 This is a structural block diagram of a charger testing system provided in an embodiment of this application. Detailed Implementation

[0020] Example 1: Refer to Figure 1 This is a flowchart illustrating a charger testing method provided in an embodiment of the present invention. The process may include at least steps S100-S400: S100: Obtain the parameter configuration set and baseline conditions, perform zero-disturbance segment acquisition and constant micro-carrier segment acquisition, and generate probe and interface configuration structure under the conditions of synchronous execution of four-wire Kelvin site registration and protocol abstract card mapping, shunt and Hall channel combination and temperature-sensitive site arrangement. S200, performs unified time base binding and event label table initialization, locks the time base by the real-time clock RTC of the microcontroller unit and broadcasts it to each channel, executes and synchronously acquires the temperature rise constraint load orchestrator of the protocol abstraction card and electronic load controller, performs probe health self-check and disturbance labeling processing, and generates the acquisition preparation package structure; S300 performs electrical transient feature extraction and ripple and overshoot fallback processing, calculates temperature gradient, interval equivalent thermal resistance and thermal inertia duration under the same time base, and performs electrothermal decoupling fusion to generate evaluation input packet structure; S400: Perform baseline-excitation-regression consistency determination and threshold comparison, anomaly three-part diagnosis and secondary verification script arrangement, uncertainty item budget and standardization report generation and configuration threshold update processing, and generate configuration and threshold update records.

[0021] S100: Obtain the parameter configuration set and baseline conditions, perform zero-disturbance segment acquisition and constant micro-carrier segment acquisition, four-wire Kelvin site registration and protocol abstract card mapping, shunt and Hall channel combination and environmental and internal temperature-sensitive site arrangement processing to obtain the probe and interface configuration structure. The S100 takes the parameter configuration set and baseline conditions as its input. The parameter configuration set consists of the charger model list, test channel number, sampling step size, threshold preset, recording strategy, and interface layout loaded during the system initialization phase. The baseline conditions are defined by the two states of the charger under test: static power-on or non-power-on, when the external excitation is disconnected, and include ambient temperature, humidity, and wind parameters, as well as the initial state of the internal heat dissipation components. Specifically, the aforementioned parameter configuration set and baseline conditions are used as input, and the system continuously executes zero-disturbance segment acquisition and constant micro-load segment acquisition. In the zero-disturbance segment acquisition, the system acquires port voltage and body temperature sensor readings at the sampling step size within a statically stable window without load switching or protocol handshake, and records the timing segment. In the constant micro-load segment acquisition, the system maintains a small constant load through a low-power electronic load, and acquires port voltage, channel current, and temperature sensor readings after stabilization to form a comparable stable segment. Furthermore, any sudden changes, jitters, out-of-bounds errors, or missing sampling frames occurring within the stable window are recorded as segment anomalies and fed back into the recording strategy entry in the parameter configuration set, triggering resampling or delayed sampling in this step until the stable segment has a usable length. Through the above processing, a stable segment set is obtained, containing the port steady-state voltage, micro-load steady-state current, and initial readings of internal and ambient temperatures. This stable segment set serves as the basis for constructing the baseline profile in subsequent mapping processes. In this step, the stable segment set does not output field names separately; instead, it is incorporated into the probe and interface configuration structure generation path along with the registration results described later.

[0022] After establishing a stable fragment set, a four-wire Kelvin site registration is performed. Four-wire Kelvin site registration refers to the process of separating power leads and measurement leads on the same electrical node, used to mark measurement sites at the output ports of the charger under test and key vias on the board under test. Port measurement sites are identified on the connector terminal side, while on-board measurement sites are identified on the via or test pad side near the power path. Registration information includes site number, site coordinates, connector type, fixture interface, and channel mapping. During site registration, the system performs a quick-start check on wire contact pressure, terminal oxidation state, and contact resistance, and writes contact state annotations into the parameter configuration set's recording strategy. If a contact abnormality is detected, a backup fixture is switched or a cleaning process is initiated before registration. This results in a site list corresponding one-to-one with the channel number; this site list is not used as an independent output field in this step but is referenced as necessary information for generating the probe and interface configuration structure.

[0023] While completing site registration, protocol abstraction card mapping is performed. The protocol abstraction card is a modular adapter component for different charging protocols, using an abstract instruction set to drive various handshake sequences. The protocol abstraction card covers general fast charging protocols and vendor-extended protocols, and provides initial definitions for common abbreviations: PD (Power Delivery) and QC (Quick Charge). Specifically, the protocol standard presets in the parameter configuration set are compared with the port steady-state identification results in the stable segment set. A matching protocol abstraction card is selected, and the abstract instruction set, handshake phase sequence, and abnormal fallback sequence are registered. In cases where multiple standards are possible, a trial order is generated according to preset priorities, and this trial order is appended to the corresponding channel mapping item in the site list. Understandably, the output of the protocol abstraction card mapping is not a separate field but is incorporated into the protocol layer description of the probe and interface configuration structure, used to call the abstract instruction set in subsequent S200 calls instead of directly manipulating the underlying handshake details.

[0024] In terms of physical measurement chain construction, a combination of shunt and Hall channels is implemented. The shunt channel acquires current readings through a small-resistance sampling device, suitable for low-frequency and steady-state current measurements; the Hall channel acquires current readings based on the magnetic field response of the Hall element, suitable for pulsating and rapidly changing scenarios. Specifically, current measurement sites in the site list are paired with available channel resources. For high dynamic range scenarios, a parallel combination of shunt and Hall channels is configured; for low-noise steady-state scenarios, a shunt main channel and a Hall backup channel are configured, and the switching threshold and priority relationship are recorded. In the combination configuration, the sampling device accuracy level, installation direction, outgoing line direction, and electromagnetic interference isolation interval are registered. When the installation direction or isolation interval is detected to be inconsistent with the boundary conditions of the recording strategy, the site is marked as requiring verification. Within this step, the site is re-verified after fine-tuning and binding correction. The verification results and correction parameters are written into the recording strategy item of the parameter configuration set. The above combination strategy and threshold are also incorporated into the current measurement substructure layer of the probe and interface configuration structure.

[0025] In constructing the thermal observation chain, environmental and internal temperature-sensitive points are arranged. Environmental temperature-sensitive points are located at representative positions in the external environment of the charger under test, recording ambient temperature and ventilation conditions. Internal temperature-sensitive points include at least three locations: the base of the heat sink, the outer surface of the casing, and the air outlet, used to observe the internal thermal path. Specifically, according to the point list and interface layout, thin-film temperature-sensitive components are attached to accessible locations on the fixture or casing, and the attachment location number, wiring path, sampling channel, and reading scale are recorded. In internal areas with air ducts or heat-conducting plates, attachments are added, and the relative distance to the heat dissipation components is recorded. Further, a short-term reading comparison is performed based on the initial temperature readings in the stable segment set, and the deviation and judgment annotations are written into the recording strategy. When situations such as loose attachment, reading drift, or sudden changes in ambient wind speed occur, the system records the abnormality annotation in this step and performs a position fine-tuning or sampling delay before reconfirmation. After completing the above arrangement, all temperature-sensitive points are incorporated into the temperature measurement substructure layer of the probe and interface configuration structure with a unified number.

[0026] During the structural integration phase, the system generates a probe and interface configuration structure based on the site list, protocol abstract card mapping results, shunt and Hall channel combination configuration, and environmental and internal temperature-sensitive site layout information. This structure includes four parts: physical layer description (ports and on-board measurement sites, four-wire Kelvin connection relationships, shunt and Hall channel mapping, wiring specifications), protocol layer description (protocol abstract card identifier, abstract instruction set, handshake sequence and fallback sequence), thermal layer description (environmental temperature-sensitive sites, internal temperature-sensitive sites, numbering and sampling channel binding), and recording strategy mirroring (contact status annotation, installation correction parameters, switching threshold and priority relationship). The index fragments of the stable fragment set are also stored in the baseline appendix area of ​​the structure. To ensure compatibility with subsequent processes, the system explicitly labels the output field name as "Probe and Interface Configuration Structure" in the structure and writes a cross-step call description within the same transaction after generation. It declares that the Probe and Interface Configuration Structure is used as input to S200 and is directly called by S210 of S200, thereby being used for unified time base binding and event label table initialization. At the same time, in the subsequent S300 and S400 stages, it is only used as a reference background and the physical and protocol description items are no longer modified.

[0027] The system establishes anomaly labeling and rollback paths in each sub-step. Anomaly labels are uniformly written into the recording strategy mirror, and rollback paths are implemented through fixture replacement, site fine-tuning, sampling delay, or backup channel switching. All correction parameters and retries are recorded in the appendix of the probe and interface configuration structure, which is used as the initial context for S210 of S200 when establishing the time base and event labeling, thereby maintaining continuity across main steps. Understandably, this step does not generate any judgment conclusion-type products; all judgment-related thresholds or consistency entries are only in a pending state in the recording strategy, waiting to be referenced in the S400 stage.

[0028] The technical effects of this step can be summarized as follows: By integrating the acquisition of zero-disturbance segments and constant microcarrier segments, the registration of four-wire Kelvin sites, the mapping of protocol abstract cards, the combination of shunt and Hall channels, and the arrangement of environmental and internal temperature-sensitive sites within the same main step, a probe and interface configuration structure that can be directly called upon by subsequent steps is formed. Anomaly annotation and correction parameters are also fixed in the structure, so that subsequent acquisition and evaluation stages have a stable physical, protocol, and thermal observation foundation.

[0029] S200, perform unified time base binding and event label table initialization, execute the load orchestrator with temperature rise constraints and synchronously acquire data, perform probe health self-check and disturbance label processing, and generate the acquisition preparation package structure; S200 uses the probe and interface configuration structure obtained from S100 as its sole external input. This configuration structure comprises four parts: a physical layer description, a protocol layer description, a thermal layer description, and a recording strategy mirror. A set of stable segments is stored in its baseline appendix area. Specifically, the aforementioned probe and interface configuration structure is loaded into the test master control unit. This unit is composed of a microcontroller unit (MCU) and a host computer. The host computer is responsible for issuing commands and aggregating data, while the MCU is responsible for local timing control and drive execution. First, a unified time base binding and event label table initialization are performed. Unified time base binding refers to establishing a unified time base across all electrical and thermal sampling channels, the protocol abstraction card handshake process, and electronic load actions. This is achieved by locking the time base on the microcontroller unit's real-time clock (RTC) and broadcasting this base to the channel scheduler, the protocol abstraction card adaptation layer, and the electronic load controller. After broadcasting, the sampling board containing the analog-to-digital converter (ADC) sends back a time base acknowledgment frame, which is recorded in the time base acknowledgment column of the event label table. Event label table initialization involves creating a structured table that grows chronologically. Each entry must include at least the timestamp, event type, event source, associated channel, associated location, handshake stage, load action, and exception flags. During initialization, system events such as time base locking, channel readiness, protocol abstraction card readiness, and load controller readiness are written first. Attention labels are added to locations marked as requiring verification in the probe and interface configuration structure. All labels are updated as the time base progresses. Understandably, once the unified time base binding is complete and the event labeling table is initialized, all driving and sampling processes will use this single time base as a reference, and the event labeling table will provide an index for subsequent data merging and exception backtracking.

[0030] After the initial construction of the time base and event table is completed, the load orchestrator under temperature rise constraints executes and synchronously acquires data. The temperature rise constrained load orchestrator is a control unit that orchestrates load commands based on dual constraints of electrical and thermal states. Its inputs are the abstract instruction set provided by the protocol abstract card, the channel mapping in the probe and interface configuration structure, the switching thresholds and priority relationships in the recording strategy mirror, and the set of stable segments in the baseline appendix area. Its outputs are the gear switching sequence and timing of the electronic load, and the handshake phase advancement and reversal commands from the protocol abstract card. Specifically, the load orchestrator first reads the handshake phase sequence from the protocol abstract card, issues handshake advancement commands in sequence, and writes events such as handshake start, handshake phase advancement, and handshake completion into the event label table. Then, it generates a progressive excitation sequence based on the stable segment set. The sequence consists of three segments: micro-step, medium-step, and scenario hybrid. Before each segment begins, a load action preview is written; after the segment begins, a load action takes effect is written; and after the action ends, a load action ends is written. All events are bound to a unified time base. Synchronous acquisition refers to the parallel sampling of voltage, current, and temperature-sensitive channels under the same time base while the load action and handshake phases are progressing. Sampling scheduling follows the channel priority relationship in the probe and interface configuration structure: in high-dynamic current scenarios, shunt and Hall effect sampling are used in parallel; in steady-state scenarios, the shunt main channel is used with Hall effect sampling reserved for backup. Voltage channels follow a dual-point sampling method, using both port and on-board points measured by a four-wire Kelvin system. Thermal channels follow the numbering order of environmental and internal temperature-sensitive points. Furthermore, temperature rise constraints are determined by the temperature gradient of the thermal channel and the temperature rise rate of adjacent sampling windows. When the temperature rise rate exceeds the gradient threshold in the recording strategy mirror, the load orchestrator writes a temperature rise over-limit warning event and reduces the load amplitude for the next time period to a preset buffer level. When the temperature rise rate is below the threshold but the temperature is close to the upper bound of the stable segment set, the load orchestrator inserts a thermal equilibrium pause event and extends the current level holding time. All insertion and rollback actions do not change the unified time base; they only generate new rows in the event annotation table. The data acquisition terminal pushes the start and end identifier of the sampling frame when each event occurs. This identifier is used as the time slice boundary in the subsequent S300 and is directly referenced without being deduced.

[0031] During load orchestration and synchronous acquisition, real-time probe health self-checks and disturbance labeling are performed. Probe health self-checks are an automatic inspection process of probe contact status and channel stability within the sampling window, including three categories: open circuit determination, short circuit determination, and reading drift determination. Open circuit determination is based on whether the reading response under a low-power probe pulse is within a minimum threshold range; short circuit determination is based on whether the differential voltage between the port and the on-board point approaches zero; and reading drift determination is based on comparing the trend offset within a continuous window with the drift threshold in the recording strategy mirror. If any determination is triggered, the system writes a probe self-check anomaly to the event labeling table, along with the point number and channel number, and simultaneously drives the channel switching logic: if a backup channel exists, it switches to the backup channel and writes "channel switching effective"; if no backup channel exists, it sets a degraded operation flag for the point and lowers its sampling priority. All switching results and degraded flags are synchronously written back to the recording strategy mirror. Disturbance labeling is the process of identifying and labeling external and internal disturbances. External disturbances include power supply jitter, sudden changes in on-site wind speed, and ambient temperature steps. Internal disturbances include electronic load quantization steps, channel sampling jitter, and protocol abstraction card handshake retries. The identification method involves searching the event labeling table for multi-source indicators near the same time scale. For example, a voltage drop without load action is labeled as power supply jitter; a periodic micro-step in current reading after load action is labeled as a quantization step. After labeling, disturbance tags are synchronously written to the data buffer for the corresponding time period. When a sudden change in wind speed occurs in the thermal channel, an environmental state change tag is added, and the current reading status of the environmental temperature-sensitive site is updated. Understandably, probe health self-checks and disturbance labeling processes run in parallel with the load orchestrator for temperature rise constraints. They coordinate through the event labeling table and a unified time base, avoiding race conditions. For all parallel processes, the host computer only acts as a recording and visual monitoring terminal and does not rewrite the real-time decision path.

[0032] After a complete progressive excitation sequence, along with its corresponding synchronous acquisition, probe health self-check, and disturbance labeling processing, is completed, the system performs a phased merging to generate an acquisition preparation package structure. This acquisition preparation package structure is an encapsulated data packet for subsequent feature extraction and evaluation. It includes at least a read-only snapshot of the event labeling table, a unified time base identifier, a channel mapping mirror, a progressive excitation sequence description, a sampling frame start-end identifier index, a channel switching and degradation flag set, a probe self-check anomaly list, a disturbance label index set, and a thermal equilibrium dwell list, along with source references pointing to the probe and interface configuration structure. Specifically, the above items are organized into a segmented structure and written to the data storage area. The label output field is named "Acquisition Preparation Package Structure," and cross-step destinations are written within the same transaction. The acquisition preparation package structure is declared as input for electrical transient feature extraction and ripple and overshoot fallback processing in S300, directly called by S310 of S300. Furthermore, the same unified time base and event labeling are used in S320 and S330 of S300, and they are not rebuilt. To maintain seamless transitions between main steps, the system simultaneously merges completed events during the event labeling table writing phase. These events reference the storage path and version number of the acquisition preparation package structure for verification by the S300 during reading.

[0033] In one embodiment, to align the channels and event streams, the test master control unit acquires the original timing data of the voltage channel, current channel, temperature-sensitive channel, and handshake event timing data, and performs alignment and sliding window slicing on a unified time base; during this process, intermediate metrics that can be used for subsequent annotation and control are formed. To enable the original multi-source sequences to have alignable time offset estimates, the following metrics are constructed: Formula① in: The optimal time offset is obtained by unifying the time base alignment; Candidates for the search offset; A timescale discretized according to a unified time base; It is a discrete sequence of port voltages; It is a discrete sequence of channel currents; : Indexing of discrete time scales Summation; Integer count index; : Returns the offset parameter that maximizes the expression within the parentheses.

[0034] The port voltage is extracted from the original timing sequence of the voltage channel and recorded as follows: Then, the channel current is extracted from the original timing sequence of the current channel and recorded as... Together, they form the relevant terms in formula ①; formula ① As an intermediate input to the S200, it provides offset estimation for the handshake-load-sampling unified time base alignment within the S200 and writes the alignment completion event to the event label table. After alignment, to form an executable sampling window mask on the event axis, the following is defined: Formula② in: For the sampling window mask, take the value range ; For rectangular window functions; Candidates for event time markers; For indicator functions; : Index of event nodes Summation; Integer count index.

[0035] Extract the phase node timestamps from the handshake event sequence and record them as follows: The action node timestamps are extracted from the load action plan and added. Set; given by a unified time base The reference alignment value. Obtained from ① Substitute ② This item projects the event nodes onto a sampling window mask under a unified time base. Its value is 1 within the event window and 0 outside the window. This is obtained via ②. This is used to guide the channel scheduler in constructing the sampling frame start and end identifiers and registering the sampling frame start and end generation event in the event label table. Combining the aforementioned processing, S200 generates two intermediate quantities in this section: time offset. With sampling window mask The interim output of this section is recorded as an alignment and mask intermediate set and passed to the next part of the load orchestration for temperature rise constraints within S200; this intermediate set is not output externally and is ultimately encapsulated into the acquisition preparation package structure.

[0036] Following the above and The load orchestrator, constrained by temperature rise, performs execution and synchronous acquisition. The load orchestrator reads the abstract instruction set of the protocol abstract card, the channel mapping in the probe and interface configuration structure, and the switching threshold and priority relationship in the recording strategy mirror, and calls the stable segment set as the starting value. Specifically, according to... The given sampling window drives the handshake process and writes the start of the handshake phase into the event label table to advance the handshake completion phase. A micro-step load is issued in the first window after the handshake is completed, a medium-step load is issued in the middle window, and a combined load is executed in the scene mixing window. Voltage and current channels are sampled simultaneously at both the port points and on-board points of the four-wire Kelvin circuit. Thermal channels are sampled simultaneously according to the numbering order of environmental and internal temperature-sensitive points. For window statistics that measure the temperature rise rate and temperature gradient in real time during acquisition, the following definition is used: Formula③ in: For window The mean temperature gradient; For window The average rate of temperature rise; For the reason The selected number A set of time-stamped samples; Window number; They are respectively sets The number of sampling points; The temperature sequence of the internal thermosensitive sites; Temperature at the internal temperature-sensitive point at time : The value; Temperature sequences of environmentally sensitive sites; : The temperature of the environmental temperature-sensitive site at time The value; The difference of the ambient temperature sequence under a unified time base discretization; : Indexing the timescale within a window within a window collection Summation; Integer count index.

[0037] The internal temperature is extracted from the original timing sequence of the temperature-sensitive channel and recorded as follows: The ambient temperature is extracted from the original time sequence of the temperature-sensitive channel and recorded as... ; obtained from ② Sure ; obtained from ③ This will be directly used to construct the constraint cost in the next equation. In temperature rise constraint control, the load level and holding time are jointly determined by the quadratic cost and the constraint threshold. A one-step rolling model prediction approach (MPC) is adopted, defined as follows: Formula④ in: For window Load level decision; Returns the candidate gear that minimizes the cost function; Candidates for decision-making; To record the set of feasible gears specified by the strategy mirror; Non-negative weights; For window The dimensional index of the response, composed of the measured port voltage and channel current; A reference vector derived from a stable set of fragments; To record the temperature rise rate threshold in the strategy mirror.

[0038] The response metrics are extracted from the original timing data of the voltage channel and the original timing data of the current channel and normalized as follows: Reference indices are extracted from the set of stable segments and denoted as follows: ; obtained from ③ Substituting the second term in ④ constitutes the penalty for deviation in temperature rise rate. The alignment and mask obtained from ① and ② implicitly define... The count and window boundary. Obtained from ④. The event labeling table records the load action as active, driving the electronic load and protocol abstraction card to execute within the same window; all channels within this window... Parallel sampling is performed on the region designated as 1. This results in a load-sampling consistent action sequence and temperature-limited acquisition data. The interim output of this section is an orchestration and acquisition intermediate set, which is then passed to the next section within S200 for probe health self-checking and disturbance labeling. The key fields of this intermediate set (handshake advancement record, load action record, window index) are ultimately incorporated into the acquisition preparation package structure for use by S300.

[0039] Following the aforementioned orchestration and acquisition intermediate set, real-time probe health self-checks and disturbance labeling are performed. Health self-checks include three types of checks: open circuit detection, short circuit detection, and reading drift detection. Short-term voltage and current responses and port-to-board differential behavior under low-power probe pulses are used to label the contact status. When any detection is triggered, the event labeling table writes a probe self-check anomaly, and the backup channel switching or degraded operation flag is updated according to the recording strategy. Disturbance labeling addresses both external and internal disturbances. External disturbances include power supply jitter and environmental state steps, while internal disturbances include electronic load quantization steps, channel sampling jitter, and protocol abstraction card handshake retries. To form a quantitative measure of disturbances in the time-frequency domain, voltage-current cross-spectral density and coherence are constructed: Formula⑤ in: For window Voltage-current coherence, range of values ; : Intra-window time-frequency index Summation; For in-window time-frequency sampling index; This refers to the Short-Time Fourier Transform (STFT). Complex conjugate operation; ; It is the square of the L2 norm, used for quadratic cost.

[0040] The port voltage channel current is extracted from the original timing sequence of the voltage channel and the original timing sequence of the current channel and denoted as follows: The result obtained from ② Give ; determined by ④ Provide the load level indicator, and Together, they form the input for judging the quantization step and power supply jitter. To fuse time-domain temperature rise limitation, frequency-domain coherence, and channel state into an operable labeled score, the following is defined: Formula⑥ in: For window The marked score; Non-negative weights; For piecewise linear mapping functions, normalize the input to ; Given by ③; This is a channel health indicator, derived from the combined logical mapping of the probe self-test abnormal channel switching and degraded operation flag.

[0041] Extract channel state events from the event label table and map them to... ; obtained from ③ The result obtained from ⑤ Substituting into ⑥, and obtaining the weighted and piecewise linear mapping, we get... .when When the disturbance threshold exceeds the threshold given by the recording strategy mirror, a specific disturbance label is written to the event label table and bound to the window index. Simultaneously, S200 attaches this label to the data segment within the window in the data buffer, for direct reference in subsequent electrical transient feature extraction and ripple and overshoot fallback processing by S300. After completing alignment, orchestration, acquisition, self-testing, and disturbance labeling, the test control unit enters the encapsulation action: generating an acquisition preparation package structure containing a read-only snapshot of the event label table, a unified time base identifier, a channel mapping mirror, a progressive excitation sequence description, a sampling frame start-end identifier index, a channel switching and degradation flag set, a probe self-test anomaly list, a disturbance label index set, and a thermal equilibrium dwell list. This output field is named "Acquisition Preparation Package Structure" and is called as input by S310 of S300. The same transaction is merged during the event label table registration phase, and the storage path and version number of the acquisition preparation package structure are recorded for verification by S300 during reading. This section summarizes the technical effects: Under a unified time base, this section generates a verifiable window-level acquisition mask, a temperature-limited load-sampling arrangement, and frequency-domain coherently driven disturbance labeling. It also encapsulates all procedural indexes and data reference paths into a structure that can be directly called by subsequent stages, forming a closed-loop record from time base, action to label.

[0042] The technical effects of this step can be summarized as follows: Through the continuous operation of unified time base binding and event label table initialization, temperature rise constraint load orchestrator execution and synchronous acquisition, probe health self-check and disturbance label processing, structured time series data and event indexes oriented towards feature extraction are formed. The acquisition preparation package structure is output and the channel status and disturbance labels are solidified under a unified time base for direct use in subsequent stages.

[0043] S300 performs electrical transient feature extraction and ripple and overshoot fallback processing, temperature gradient and equivalent thermal resistance and thermal inertia duration calculation, electrothermal decoupling and fusion processing, and generates evaluation input packet structure; The S300 uses the acquisition preparation package structure output by the S200 as its sole input source. This acquisition preparation package structure includes a read-only snapshot of the event labeling table, a unified time base identifier, a channel mapping mirror, a progressive excitation sequence description, a sampling frame start-end identifier index, a channel switching and degradation flag set, a probe self-check anomaly list, a disturbance tag index set, and a thermal equilibrium dwell list. Specifically, the aforementioned acquisition preparation package structure is loaded into the feature processing unit. The feature processing unit consists of a data buffer, a timing aligner, and a feature generator, which establishes a mapping relationship between the time scale and the sampling frame based on the unified time base identifier during the loading phase. First, the sampling frame start-end identifier index is traversed, and the three types of load actions in the progressive excitation sequence—micro-step, medium-step, and scene mixing—are positioned as event windows on the time axis. Then, the load action prediction, load action activation, load action end, thermal equilibrium dwell, and handshake phase advancement in the event labeling table are mapped to the same set of time boundaries. Furthermore, the disturbance label index set is parsed, and the time period marked by the handshake retry of the sampling jitter protocol abstract card of the power supply jitter environment state change quantification step channel is marked as the candidate elimination area; understandably, the candidate elimination area is regarded as the reference area during feature generation, and the threshold rule determines whether it participates in the statistics or is replaced by the mirror statistical value of the same window.

[0044] In the process of extracting electrical transient features and handling ripple and overshoot / fallback, electrical transient features refer to the rapid response parameters generated by the voltage and current channels after load action triggering, including rise time, overshoot amplitude, fallback time, short-time voltage drop amplitude step, and differential point deviation. Ripple refers to the periodic fluctuation statistics within the steady-state window, including peak-to-peak values ​​and amplitude indicators corresponding to the main frequency. Overshoot / fallback handling is a comprehensive process for locating the peak values ​​of voltage and current, identifying the fallback interval, and determining the end of fallback after load action. Specifically, each micro-step and mid-step window is moved forward by one preparatory segment as a baseline reference segment. Within the baseline reference segment, the differential voltage and differential current between the port point and the on-board point are calculated. When entering the action segment, the final reading of the baseline reference segment is used as the zero-point reference to complete the location of rise time and short-time voltage drop. For overshoot and fallback identification, the overshoot peak position is determined by combining peak search and threshold determination. Then, the fallback end point is determined based on the fallback termination threshold and fallback slope threshold, and the time span from the end of the action segment to the end of the fallback is recorded. Ripple statistics are performed during the thermal equilibrium pause or the semi-steady-state interval after the action segment. The steady-state interval is identified by the combination of the end of the load action and the thermal equilibrium pause in the event labeling table. For segments in the candidate rejection area, a mirror replacement strategy is adopted, with the mirror segment sourced from the adjacent steady-state window of the same type of action. When the mirror segment is unavailable, the ripple index is recorded as null and a steady-state segment insufficiency warning is written to the alarm queue of the feature generator. At the same time, a constrained fusion flag is attached to the entry. The processing of channel switching and degradation flag sets follows the principle of continuous splicing and source labeling. At the switching point, the source channel label is attached to the feature entry so that the splicing boundary can be identified in the subsequent fusion stage.

[0045] In the calculation of temperature gradient, equivalent thermal resistance, and thermal inertia duration, the temperature gradient is defined as the temperature difference between the internal temperature-sensitive site and the environmental temperature-sensitive site at the same time scale, and the spatial difference sequence between multiple sites within the machine. Equivalent thermal resistance is defined as a structured comparison index established between the power consumption step triggered by load action and the temperature response. Thermal inertia duration is defined as the time span from the start of the action to entering the stable bandwidth. Specifically, based on the thermal balance dwell list and the progressive excitation sequence description, a thermal response window is established at the beginning of each action segment, and synchronous readings of the internal and environmental temperature-sensitive sites are collected within the window. The instantaneous temperature gradient between the computer and the environment is also recorded, and a spatial difference sequence is formed between multiple sites within the machine for subsequent path determination. For the calculation of equivalent thermal resistance, a segmented comparison method is adopted without introducing formulas: during the period from the start of the action segment to the characteristic time point, the rising amplitude of the temperature gradient is paired one-to-one with the power consumption step label defined for the action segment. The comparison result of the amplitude pair is recorded as a structured equivalent term, and several equivalent terms are summarized in multiple excitation segments to form an interval equivalent thermal resistance description. For thermal inertia duration, upper and lower bounds of stable bandwidth are established in the temperature gradient time series. Using the start point of thermal equilibrium dwell as a reference point, the time point of first entry and sustained dwell is searched, and the difference between the two corresponds to the thermal inertia duration entry. If a change in environmental state causes stable bandwidth drift, stable bandwidth records before and after the change are retrieved from the event annotation table. The bandwidth record closest to the action segment is used; if no match can be found, a constraint fusion flag is added to the entry, and the original dwell time is retained. The above three types of thermal entries and electrical transient entries use the same unified time base identifier, and a channel source and location number are added to each record to ensure traceability of cross-domain references.

[0046] In the electrothermal decoupling fusion process, electrothermal decoupling fusion is defined as the process of window alignment, disturbance removal, and semantic binding of electrical transient entries and thermal entries on a unified time base, outputting a comprehensive feature carrier that can be directly called in the evaluation stage. Specifically, firstly, based on the sampling frame start-end identifier index, a fusion window is established for each load action segment. The left boundary of the window is the end point of the load action forecast, and the right boundary is the start point of the steady-state segment from the end of the load action to the thermal equilibrium dwell time. Inside the window, the removal rules are executed according to the disturbance label index of the event label table: segments with power supply jitter and protocol abstract card handshake retry are removed from electrical entries, and segments with environmental state changes are removed from thermal entries. If the window length is insufficient to form an entry after removal, cross-segment supplementation is performed for similar action segments, with the priority order of similarity being the same load amplitude, adjacent same type, and adjacent different type. Subsequently, semantic binding was performed on the retained electrical and thermal entries: rise time was bound to the initial slope of short-term voltage drop and temperature gradient; overshoot amplitude and fallback time were bound to the dwell position of thermal inertia duration; and steady-state ripple was bound to the interval equivalent term of equivalent thermal resistance. The binding relationship was represented by the entry number and event window number, without introducing variable notation. To handle splicing boundaries caused by channel switching, segments with consistent sources were prioritized for binding during fusion. In cases where the sources were inconsistent and backtracking was not possible, cross-source binding flags were added to the entries, and references were prompted according to entry weights during the evaluation phase. After fusion, available intervals and confidence labels were established for all entries. Available intervals came from the time coverage of the window, and the proportion of confidence labels from the candidate rejection area was consistent with the entry source. At the same time, entries with constrained fusion flags were included in a separate index for broadband reference during threshold comparison in the evaluation phase.

[0047] Regarding anomaly handling and recording, if a persistent zero value is found in the difference between port sites and on-board sites during electrical transient feature extraction, and this value is not within the range of degraded operation flags, it is determined to be a site access anomaly entry. A site access verification event is added to the event annotation table, and the entry is marked as restricted read-only, providing only time location information and not participating in fusion. If the temperature gradient direction is found to be consistently opposite to the spatial difference sequence direction during thermal calculations, and there is no change in environmental conditions, it is recorded as a duct anomaly entry and moved to the alarm queue. If the proportion of rejections exceeds the threshold within the fusion window, the window is automatically merged into an adjacent window of the same type, and a window merging annotation is added to the fusion result so that changes in time coverage can be known during the evaluation phase. All the above anomalies and annotations are written into the entry extension field. The entry extension field and the event annotation table are associated through window numbers, preserving both the data itself and process information.

[0048] After feature extraction, thermal calculation, and decoupling fusion are completed, the system enters the encapsulation stage, generating the evaluation input packet structure. The evaluation input packet structure consists of four parts: first, an electrical feature input set, including rise time, overshoot amplitude, fall time, short-time voltage drop, and steady-state ripple, along with their sources and available ranges; second, a thermal feature set, including temperature gradient, interval equivalent thermal resistance, and thermal inertia duration, along with their sources and available ranges; third, a fusion index and confidence label, including window number, entry binding relationship, cross-source binding flag, unconstrained fusion flag, and confidence level; and fourth, a reference guide, including reference paths to channel switching and degradation flag sets, lookup paths to the event label table, and index paths to the sampling frame start-end identifier index. Understandably, the output field of the evaluation input packet structure is named "Evaluation Input Packet Structure," and during generation, it is written with cross-step destinations, declaring direct calls from the S400 baseline-excitation-regression consistency judgment and threshold comparison. Within the same main step, the entry extension field is read-only protected to prevent rewriting of the original features and fusion records during the evaluation stage. To maintain the context formed by S100 and S200, a read-only reference to the probe and interface configuration structure and the version number of the acquisition preparation package structure are attached to the meta-information area of ​​the evaluation input packet structure for subsequent backtracking.

[0049] In summary, the technical effects of this step are as follows: by segmenting and extracting electrical transient characteristics and steady-state ripple under a unified time base, windowing calculation of temperature gradient and interval equivalent thermal resistance and thermal inertia duration, and decoupling, fusion and labeling encapsulation for disturbances and channel switching, an evaluation input packet structure with time continuity and traceability of origin is formed.

[0050] S400, perform baseline-excitation-regression consistency determination and threshold comparison, anomaly three-part diagnosis and secondary verification script arrangement, uncertainty item budget and standardized report generation and configuration threshold update processing, and generate configuration and threshold update records; S400 uses the evaluation input package structure generated by S300 as its sole input. This structure includes an electrical feature input set, a thermal feature set, a fusion index and confidence markers, and a reference guide. It also read-only references the event label table and unified time base identifier generated by S200 in the metadata area, while retaining background references to the probe and interface configuration structure of S100. Specifically, the aforementioned evaluation input package structure is loaded into the decision and archiving unit. This unit reads the fusion index window by window according to the unified time base, calling up the rise time, overshoot amplitude, fall time, short-time voltage drop, and steady-state ripple entries from the electrical feature input set. Simultaneously, it calls up the temperature gradient, interval equivalent thermal resistance, and thermal inertia duration entries from the thermal feature set. Based on the window number and entry binding relationship, a three-segment correspondence is established: baseline, excitation, and regression. The baseline segment comes from the baseline appendix formed by the zero-disturbance segment acquisition and constant micro-load segment acquisition of S100. The excitation segment corresponds to the progressive excitation sequence formed by the load orchestrator execution under S200 temperature rise constraints. The regression segment corresponds to the re-acquisition interval after the excitation ends. The decision and archiving unit first performs baseline-excitation-regression consistency determination and threshold comparison. Using the port steady-state voltage, micro-load steady-state current, and initial temperature readings of the baseline segment as references, it compares the window stability of rise time and short-time voltage drop within the excitation segment. Within the regression segment, it compares whether steady-state ripple and temperature gradient have regressed to the baseline's allowable bandwidth. For each comparison result, it consults the threshold path in the reference guide and reads the current threshold list for threshold comparison. The threshold list is derived from the current entries in the parameter configuration set. During the comparison process, the decision and archiving unit only generates decision labels and deviation magnitudes, without rewriting the threshold list. When a fusion index-labeled entry is to be constrained for fusion, a temporary comparison is performed using the broadband reference given in the reference guide, and a broadband reference label is added to the entry and stored in the decision cache.

[0051] After consistency and threshold comparison are completed, the judgment and archiving unit executes the anomaly triad diagnosis and secondary verification script arrangement. The anomaly triad diagnosis is a process of classifying anomaly candidates into three categories: device-endogenous, external disturbance, and test link. Device-endogenous anomalies include ripple amplification, thermal inertia anomalies, and overshoot / fallback delay. External disturbances include power supply jitter and environmental state changes. Test link anomalies include probe contact instability, high-frequency channel switching, and site access anomalies. Specifically, the judgment and archiving unit cross-references the temporal neighborhood of anomaly candidate entries based on the fusion index and event label table. If there is no recorded load action before or after the anomaly occurs but the voltage drops suddenly, it is classified as an external disturbance. If the anomaly occurs after the load action and the thermal inertia duration significantly expands while the environmental state remains unchanged, it is classified as a device-endogenous anomaly. If the anomaly coincides with a channel switching or degradation flag, it is classified as a test link. After classification, secondary verification scripts are compiled for each type of anomaly. These scripts consist of action instructions and acquisition instructions. Action instructions include temporarily changing the load amplitude, inserting a thermal equilibrium pause, and triggering a handshake back check with the protocol abstraction card. Acquisition instructions include repeating sampling within a specific window, locking a backup channel for read-only sampling, and adding short-term dense sampling at temperature-sensitive sites. To prevent rewriting of the original data, all secondary verification scripts are recorded as a script list. This list is not directly executed within the main step; it is only written into the output product and a reference path is provided. When an anomaly is classified as a test link and the site access anomaly is marked as persistent, a fixture verification suggestion entry is added and bound to the site number in the probe and interface configuration structure for subsequent site verification upon returning to S100.

[0052] After the secondary verification script is compiled, the decision and archiving unit performs uncertainty item budgeting, standardized report generation, and threshold update processing. Uncertainty item budgeting is a process of source analysis and interval quantization for four types of error sources: electrical, thermal, load, and sampling. Source analysis is based on the source channel, site number, and splicing boundary recorded in the event labeling table and channel mapping mirror record. Interval quantization is based on window coverage and candidate rejection ratio. The budget results form item budget entries, which are associated with the entry number and window number, and are marked with source consistency and rejection ratio. Standardized report generation is a process of structuring the consistency judgment results, threshold comparison conclusions, anomaly three-part diagnostic list, secondary verification script list, and item budget entries. The report structure includes four sections: judgment overview, window index, entry details, and traceability path. The traceability path provides a lookup path for the S200 event labeling table and unified time base identifier, as well as a reference path for the S100 probe and interface configuration structure and the S300 evaluation input package structure. The threshold update process involves mapping the deviation statistics and broadband reference usage annotations generated during this judgment process into threshold candidate update items according to the threshold update strategy. The threshold candidate update items do not directly overwrite the current threshold list, but coexist with the threshold list and record the effective conditions and evaluation batch number. The effective conditions include context such as equipment model, protocol type and environmental classification. When there are consistent deviations across multiple windows in this batch and the anomalies are not classified into the external disturbance category or the test link category, the threshold candidate update items are marked with priority review to prompt manual or system review during the parameter configuration set loading stage of S100.

[0053] To maintain seamless workflow across main processes, the decision and archiving unit, after completing budget, reporting, and threshold processing, enters the encapsulation phase, generating configuration and threshold update records. These records comprise four fields: decision, anomaly, budget, and threshold. The decision field records the labels, deviation magnitudes, and bandwidth references for baseline-incentive-regression consistency determination and threshold comparison; the anomaly field records the three-part anomaly diagnostic results, a list of secondary verification scripts, and fixture review recommendations; the budget field records the consistency of budget items with their sources, the percentage of items removed, and window coverage; and the threshold field records candidate threshold updates, their effective conditions, and evaluation batch numbers, including a mapping to the current threshold list. Upon completion of encapsulation, the cross-step destination is written within the judgment and archiving unit, declaring that the configuration and threshold update records are read by S110 of S100 as input to the parameter configuration set in S100. When loading the parameter configuration set, S100 simultaneously loads the threshold domain and fixture verification suggestions of this record for use as a preliminary reference for the next batch of zero-disturbance fragment acquisition, constant micro-load fragment acquisition, and site registration. Simultaneously, a read-only reference path is written into the report structure for external quality traceability systems to access. Understandably, the configuration and threshold update records are the final product within this main step and are archived according to a unified time base and batch number to support continuous access by the testing system in production lines and R&D stages.

[0054] Regarding anomaly and boundary handling, if a regression segment is missing or time coverage is insufficient during the consistency determination stage, the determination and archiving unit writes a regression segment insufficiency annotation in the determination field and automatically generates a regression supplementation secondary verification script in the anomaly field. If broadband references are used extensively during the threshold comparison stage, a broadband reference high percentage annotation is written in the threshold field, and a review must-see mark is added to the corresponding window's sub-budget items. If source consistency is not met for an extended period during the budget stage, the determination and archiving unit adds a source inconsistency review item to the anomaly field and binds the relevant channel switching event and location number. All the above annotations and scripts are written into the configuration and threshold update records during encapsulation and do not affect the read-only integrity of the report structure for this batch.

[0055] Regarding the report generation strategy, the standardized report generation unit follows a fixed paragraph template and a dynamic appendix mechanism. The overall judgment overview is placed at the beginning, the window index and item details are expanded chronologically, and the traceability path and sub-budgets are placed in the appendix area. When a secondary verification script list exists, a script reference page is generated at the end of the report. This page only includes the script number, triggering conditions, and execution interface description, excluding the execution results. When a fixture review recommendation exists, it is indicated on the report cover, and a link description pointing to the probe and interface configuration structure is generated in the index. After the report is generated, the judgment and archiving unit adds a report archiving completion event to the event annotation table and writes the report path and version number into the configuration and threshold update records for subsequent review.

[0056] Understandably, the entire processing chain of S400 does not rewrite the event labeling table of S200, the evaluation input packet structure of S300, or the probe and interface configuration structure of S100. All new content is written to the configuration and threshold update record as the sole target, completed within the same transaction, and archived in batches. The output field of the configuration and threshold update record is named "Configuration and Threshold Update Record". This output takes effect within this main step and is read in the next batch of S100, forming the update source of the parameter configuration set and the entry point for fixture verification, thereby completing the closed-loop chain of baseline-acquisition-evaluation-archiving.

[0057] In summary, the technical effects of this step are as follows: By completing the consistency determination and threshold comparison of baseline-excitation-regression under a unified time base, the script arrangement of anomaly three-part diagnosis and secondary verification, the budgeting of partial uncertainty and the generation of standardized reports, and the encapsulation of update suggestions into rechargeable configuration and threshold update records, the parameters and fixtures for the next batch of verification input are formed, realizing the closed-loop operation and continuous convergence of the charger test system.

[0058] Example 2: Figure 2 A structural block diagram of a charger testing system according to an embodiment of the present invention is shown. Figure 2 As shown, the structure may include: The probe and interface configuration unit 01 is used to perform zero-disturbance segment acquisition and constant micro-load segment acquisition, four-wire Kelvin site registration, protocol abstract card mapping, shunt and Hall channel combination, and environmental and internal temperature-sensitive site arrangement processing according to the parameter configuration set and baseline conditions, and generate the probe and interface configuration structure. Specifically, it receives input from the parameter configuration set and baseline conditions, completes continuous execution of port steady-state sampling and micro-load steady-state sampling under the constraints of fixture and channel resources, registers the number and coordinates of port sites and in-board sites and establishes channel mapping, completes protocol abstract card selection and handshake phase sequence registration, arranges the master and backup relationship and switching threshold of shunt channels and Hall channels, pastes and numbers environmental and internal temperature-sensitive sites and records sampling channels and reading scales; sets abnormal markings and backoff paths for contact instability, reading drift and wiring interference, corrects parameter writing record strategy mirroring; outputs the probe and interface configuration structure and attaches a stable segment set in the baseline appendix area, hands the probe and interface configuration structure over to the time base and event marking unit for loading, and registers cross-step call instructions in the storage area for subsequent unit review.

[0059] The Time Base and Event Labeling Unit 02 is used to load the probe and interface configuration structure, complete the unified time base binding and event labeling table initialization in the test main control unit, and establish an event structure that grows in chronological order. Specifically, it receives the probe and interface configuration structure from the probe and interface configuration unit, broadcasts a single time base to the acquisition channel and load control end according to the registration results of the channel scheduling and protocol adaptation layer, and collects the time base confirmation frame to generate time base lock entries; it establishes an event labeling table and writes system events such as channel ready, abstract card ready, and load control ready, adds attention labels to the points that need to be verified and updates them over time; it triggers time base replay and channel reset for broadcast anomalies and confirmation missing, and writes the relevant labels into the event structure; it outputs an event labeling table with a unified time base and a time base identifier for parallel use by the load orchestration and synchronous acquisition unit and the probe health and disturbance judgment unit, and maintains an event cursor in the buffer for subsequent reading.

[0060] The load orchestration and synchronous acquisition unit 03 is used to perform load orchestration and synchronous acquisition processing of progressive excitation sequences according to the protocol abstract card instruction set and temperature rise constraints, and record load actions and sampling events under a unified time base. Specifically, it receives the event label table and time base identifier from the time base and event label unit, as well as the channel mapping and recording strategy mirror from the probe and interface configuration unit. It advances the handshake stage according to the abstract instruction set and generates a progressive excitation sequence of micro-step, medium-step and scene mixture. The start and end of the action are written into the event entries according to the unified time base. Parallel sampling is performed on the voltage channel and current channel within the action window, and numbered sequential sampling is performed on the temperature-sensitive channel. The start and end identifiers of the sampling frames are aligned with the load action entries. When the temperature rise trend approaches the threshold, a thermal equilibrium dwell event is inserted and the subsequent level amplitude is adjusted. The relevant entries are written into the event structure. The phased sampling buffer and load action record are output, and the start and end identifiers of the sampling frames are passed to the probe health and disturbance judgment unit and the feature extraction and fusion unit for reference.

[0061] The probe health and disturbance determination unit 04 is used to perform probe health self-checks and disturbance labeling during the acquisition process. It identifies and labels the probe contact status, channel stability, and external and internal disturbances. Specifically, it receives the phased sampling buffer and load action records from the load orchestration and synchronous acquisition unit, as well as the event labeling table and time base identifier from the time base and event labeling unit. It performs inspections on open circuits, short circuits, and reading drifts and writes self-check entries with site numbers and channel numbers into the event structure. It triggers primary / backup channel switching or downgraded sampling and synchronously registers the switching effective entries. It identifies disturbances such as power supply jitter, sudden changes in environmental state, quantization steps, and handshake retries and writes disturbance labels for the corresponding time periods. After completing the phase merging, it encapsulates the event structure snapshot, channel mapping mirror, sampling frame start and end identifier index, and switching flag set into an acquisition preparation package structure and outputs it to the feature extraction and fusion unit for calling. At the same time, it retains the abnormal list for cross-retrieval by the consistency determination and reporting unit.

[0062] Feature extraction and fusion unit 05 is used to perform electrical transient feature extraction, ripple and overshoot fallback processing, temperature gradient and equivalent thermal resistance and thermal inertia duration calculation, and electrothermal decoupling fusion processing on the sampled data of voltage channel, current channel, and temperature-sensitive channel based on the acquisition preparation packet structure, and generate an evaluation input packet structure. Specifically, it receives the acquisition preparation packet structure from the probe health and disturbance judgment unit, locates the action window according to the sampling frame start and end identifier index, performs rejection and mirror replacement on the candidate rejection area, and extracts rise time, overshoot amplitude, fallback time, short-time voltage drop, and steady-state ripple, etc. The system calculates temperature gradient, interval equivalent thermal resistance, and thermal inertia duration entries based on temperature-sensitive point readings, and extends the annotation of duct anomalies and bandwidth drift records. Under a unified time reference, it completes window alignment and semantic binding of electrothermal entries, establishes the correspondence between entry numbers and window numbers and source annotations, and writes restricted read-only identifiers for unavailable entry scenarios. It outputs the evaluation input packet structure and provides fusion index and confidence annotation, which are read by the consistency judgment and reporting unit. At the same time, it retains reference guidance and channel switching associations in the storage area for subsequent review.

[0063] The Consistency Judgment and Reporting Unit 06 is used to call the evaluation input package structure, perform baseline-excitation-regression consistency judgment and threshold comparison, anomaly three-part diagnosis and secondary verification script arrangement, uncertainty item budgeting and standardized report generation, and generate configuration and threshold update records. Specifically, it receives the evaluation input package structure and fusion index from the feature extraction and fusion unit, and references the event labeling table and time base identifier provided by the time base and event labeling unit, as well as the baseline appendix entries provided by the probe and interface configuration unit. It completes the three-segment correspondence and threshold comparison in the window dimension, and generates judgment labels and deviation magnitudes. According to the channel switching flag, disturbance label and time neighborhood, the anomaly candidates are classified into equipment endogenous type, external disturbance type and test link type, and secondary verification scripts and fixture review suggestions are arranged. Item budgeting is performed on electrical, thermal, load and sampling sources and source consistency and elimination ratio labeling are established. The judgment overview, window index and traceability path are assembled to form a standardized report. The above entries are encapsulated as configuration and threshold update records and output to the threshold and parameter update unit. At the same time, the report path and version number are recorded in the archive area for the starting unit to read.

[0064] The threshold and parameter update unit 07 is used to write the threshold fields and fixture review suggestions from the configuration and threshold update records into the parameter configuration set, providing loading input for the next batch of tests and realizing the closed loop of the testing process. Specifically, it receives the configuration and threshold update records from the consistency judgment and reporting unit, establishes a coexistence relationship between the candidate threshold update items and the current threshold list, and registers the effective conditions and evaluation batch number. It also establishes a binding index for the fixture review suggestions and the site number. The threshold fields are written into the threshold list entries of the parameter configuration set, and the fixture review suggestions are written into the review entry entries of the record strategy image, and the loading order and version number are updated. After writing is completed, the updated parameter configuration set and record strategy image are sent back to the probe and interface configuration unit as the input source for the next batch. At the same time, the closed loop completion event and timestamp of this batch are registered in the archive area to maintain the continuity of the cross-batch traceability link.

Claims

1. A charger testing method, characterized in that, include: S100: Obtain parameter configuration set and baseline conditions, perform zero-disturbance segment acquisition and constant microcarrier segment acquisition, four-wire Kelvin site registration, protocol abstract card mapping, shunt and Hall channel combination and temperature-sensitive site arrangement, and generate probe and interface configuration structure. The parameter configuration set includes a list of charger models, test channel numbers, sampling step size, threshold preset, recording strategy, and interface layout; the baseline conditions are defined by two states of the charger under test when the external excitation is disconnected: stationary power-on or non-power-on, and include ambient temperature, humidity, and wind parameters and the initial state of the internal heat dissipation components. The zero-disturbance segment acquisition and constant micro-load segment acquisition include: acquiring port voltage and body temperature sensor readings in a statically stable window with no load switching and no protocol handshake, according to the sampling step size; maintaining a small constant load through a low-power electronic load, and acquiring port voltage, channel current, and temperature sensor readings after stabilization; recording any sudden changes, jitter, out-of-bounds, or missing sampling frames that occur within the window as segment anomalies and triggering repeated sampling or delayed sampling until a stable segment set with usable length is obtained; the stable segment set includes the port steady-state voltage, micro-load steady-state current, and initial readings of internal and ambient temperatures; The protocol abstract card mapping includes: comparing the preset protocol type in the parameter configuration set with the port steady-state identification results in the stable segment set, selecting the matching protocol abstract card, and registering the abstract instruction set, handshake phase sequence, and abnormal rollback sequence; generating an attempt order according to the preset priority, and attaching the attempt order to the channel mapping item corresponding to the site list; The probe and interface configuration structure includes: a physical layer description, a protocol layer description, a thermal layer description, and a recording strategy mirror. The physical layer description includes the port and on-board measurement points, the four-wire Kelvin connection relationship, the shunt and Hall channel mapping, and the wiring specifications. The protocol layer description includes the protocol abstract card identifier, the abstract instruction set, the handshake sequence, and the fallback sequence. The thermal layer description includes the environmental temperature-sensitive points, the internal temperature-sensitive points, and the binding relationship between the numbering and the sampling channel. The recording strategy mirror includes contact status annotations, installation correction parameters, switching thresholds, and priority relationships. The probe and interface configuration structure also stores a set of stable segments in its baseline appendix area. S200. Based on the probe and interface configuration structure, a unified time base binding and event labeling table initialization are performed. The real-time clock (RTC) of the microcontroller unit locks the time base and broadcasts it to each channel, protocol abstraction card, and electronic load controller. The load orchestrator with temperature rise constraints is executed and synchronously acquired. The probe health self-check and disturbance labeling are processed to generate the acquisition preparation package structure. The event label table includes fields for time stamp, event type, event source, associated channel, handshake phase, load action and exception flag, and has been written with system events for time base lock, channel ready, protocol abstract card ready and load controller ready. The temperature rise constraint includes determining the load level based on the temperature gradient and the temperature rise rate, reducing the load amplitude when the temperature rise rate exceeds the threshold, and inserting a thermal equilibrium dwell event when the temperature approaches the upper limit. The acquisition preparation package structure includes: a read-only snapshot of the event labeling table, a unified time base identifier, a channel mapping mirror, a progressive excitation sequence description, a sampling frame start and end identifier index, a channel switching and degradation flag set, a probe self-test anomaly list, a disturbance tag index set, and a thermal equilibrium dwell list. The acquisition preparation package structure includes a source reference pointing to the probe and interface configuration structure; S300. Based on the acquisition preparation packet structure, perform electrical transient feature extraction and ripple and overshoot fallback processing, calculate temperature gradient, interval equivalent thermal resistance and thermal inertia duration under the same time base, and perform electrothermal decoupling fusion to generate evaluation input packet structure. The evaluation input package structure includes: an electrical feature input set, a thermal feature set, a fusion index and confidence annotations, and a reference guide; The electrical feature input set includes rise time, overshoot amplitude, fall time, short-time voltage drop, and steady-state ripple entries; The thermal feature set includes entries for temperature gradient, interval equivalent thermal resistance, and thermal inertia duration. The fusion index and confidence label include window number, entry binding relationship, cross-source binding flag, fusion flag to be constrained, and confidence level; The reference guidance includes a reference path for the channel switching and degradation flag set, a lookup path for the event label table, and an index path for the sampling frame start and end identifier index. S400. Based on the evaluation input package structure, perform baseline-excitation-regression consistency determination and threshold comparison, anomaly three-part diagnosis and secondary verification script arrangement, uncertainty item budget and standardized report generation, and configuration threshold update processing to generate configuration and threshold update records; the anomaly three-part diagnosis is to classify anomaly candidates into equipment-endogenous, external disturbance and test link categories. The configuration and threshold update record includes: a decision domain, an anomaly domain, a budget domain, and a threshold domain; The decision domain records the labels, deviation magnitude, and broadband reference annotations used for the baseline-excitation-regression consistency determination and threshold comparison. The anomaly field records the three-part diagnostic results of anomalies, the list of secondary verification scripts, and the suggestions for fixture review. The budget field records the consistency of the sub-budget items and their sources, the percentage of items removed, and the window coverage. The threshold field records the threshold candidate update items, their effective conditions, and the evaluation batch number, and includes a mapping relationship with the current threshold list.

2. The method according to claim 1, characterized in that, The process of binding a unified time base includes: Unified time base binding refers to the process of establishing a unified time base among all electrical and thermal sampling channels, protocol abstraction card handshake process, and electronic load actions. The time base is locked by the real-time clock (RTC) of the microcontroller unit and broadcast to the channel scheduler, protocol abstraction card adaptation layer, and electronic load controller. After the broadcast is completed, the sampling board where the analog-to-digital converter (ADC) is located sends back a time base confirmation frame, which is recorded in the time base confirmation column of the event label table.

3. The method according to claim 1, characterized in that, The load orchestrator with temperature rise constraints is a control unit that orchestrates load commands based on dual constraints of electrical and thermal states. Its inputs include the abstract instruction set, channel mapping, recording strategy image, and stable segment set provided by the protocol abstraction card, and its outputs are electronic load gear switching sequences and protocol handshake advancement commands.

4. The method according to claim 1, characterized in that, The synchronous acquisition process includes: Synchronous acquisition involves parallel sampling of the voltage channel, current channel, and temperature-sensitive channel under the same time base while the load action and handshake phases are progressing. The sampling scheduling follows the channel priority relationship in the probe and interface configuration structure, realizing current shunting and Hall parallel sampling, voltage Kelvin dual-point parallel sampling, and thermal channel number sequential sampling.

5. The method according to claim 1, characterized in that, The determination of temperature rise constraint is driven by the temperature gradient of the thermal channel and the temperature rise rate of adjacent sampling windows. When the temperature rise rate exceeds the gradient threshold in the recording strategy mirror, the load orchestrator writes a temperature rise over-limit warning event and reduces the load amplitude. When the temperature approaches the upper limit of the stable segment set, a thermal equilibrium dwell event is inserted. The upper limit of the stable segment set refers to the upper limit of the temperature reading in the stable segment set.

6. The method according to claim 1, characterized in that, The probe health self-test performs three types of inspections within the sampling window: open circuit judgment, short circuit judgment, and reading drift judgment. These are based on the low-power probe pulse response, port differential voltage, and continuous window trend offset, respectively. When an abnormality is detected, a probe self-test abnormality is written and the channel switching logic is driven.

7. The method according to claim 6, characterized in that, The channel switching logic switches to the backup channel when a backup channel exists and writes the channel switching to take effect. When no backup channel exists, it sets a downgrade operation flag for the site and reduces its sampling priority. All results are synchronously written back to the recording strategy mirror.

8. The method according to claim 1, characterized in that, The disturbance labeling process identifies and labels external and internal disturbances. External disturbances include power supply jitter, sudden changes in on-site wind speed, and ambient temperature steps. Internal disturbances include electronic load quantization steps, channel sampling jitter, and protocol abstraction card handshake retries. The corresponding disturbance labels are written into the event labeling table.

9. A charger testing system for performing the method of any one of claims 1-8, characterized in that, include: The probe and interface configuration unit is used to perform zero-disturbance segment acquisition and constant microcarrier segment acquisition, four-wire Kelvin site registration, protocol abstract card mapping, shunt and Hall channel combination, and environmental and internal temperature-sensitive site arrangement processing according to the parameter configuration set and baseline conditions, and generate the probe and interface configuration structure. The time base and event labeling unit is used to load the probe and interface configuration structure, complete the unified time base binding and event labeling table initialization in the test main control unit, and establish an event structure that grows in chronological order. The load orchestration and synchronous acquisition unit is used to perform load orchestration and synchronous acquisition processing of progressive excitation sequence according to the protocol abstract card instruction set and temperature rise constraint conditions, and record load actions and sampling events under a unified time base. The probe health and disturbance determination unit is used to perform probe health self-check and disturbance labeling processing during the acquisition process, and to identify and label the probe contact status, channel stability and external and internal disturbances; The feature extraction and fusion unit is used to perform electrical transient feature extraction, ripple and overshoot fallback processing, temperature gradient and equivalent thermal resistance and thermal inertia duration calculation, and electrothermal decoupling fusion processing on the sampled data of voltage channel, current channel and temperature-sensitive channel based on the acquisition preparation packet structure, and generate evaluation input packet structure. The consistency determination and reporting unit is used to call the evaluation input package structure, perform baseline-excitation-regression consistency determination and threshold comparison, anomaly three-part diagnosis and secondary verification script orchestration, uncertainty item budgeting and standardized report generation, and generate configuration and threshold update records; The threshold and parameter update unit is used to write the threshold domain and fixture review suggestions of the configuration and threshold update record into the parameter configuration set, providing loading input for the next batch of tests and realizing the closed loop of the test process.

Images