CN-121984153-A - Self-adaptive low-power-consumption dormancy control method for secondary fusion capacitor power taking

CN121984153ACN 121984153 ACN121984153 ACN 121984153ACN-121984153-A

Abstract

The invention discloses a self-adaptive low-power consumption dormancy control method for secondary fusion capacitor electricity taking, which comprises the steps of generating an energy load observation record by collecting a capacitor end voltage measurement value, a temperature measurement value and a load event count value, recursively updating equivalent leakage parameters and load impact coefficients of a capacitor to obtain a prediction parameter vector, generating a self-learning safety margin value according to the prediction parameter vector, further extrapolating a prediction window voltage lower bound to obtain an available energy prediction value, mapping an operation trigger event into an operation mark to be executed to look up a table to obtain a task energy budget value, completing dormancy/awakening decision according to a prediction gating criterion of the available energy prediction value, the task energy budget value and the self-learning safety margin value, checking through consistency, restraining false awakening, reducing undervoltage reset probability and improving stability under a working condition abrupt change scene.

Inventors

KONG DONGBO
GUO KAI
WANG FEIFEI
WEN JIANQIANG
TANG GUOPING
TAN MINGYI

Assignees

江苏丹通电气有限公司

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (10)

1. A self-adaptive low-power-consumption dormancy control method for secondary fusion capacitor electricity taking is characterized by comprising the following steps: In each preset sampling period, collecting a capacitance end voltage measured value, a temperature measured value and a load event count value of a secondary fused capacitance power taking loop, writing a sampling period index, and generating an energy load observation record; Updating the equivalent capacitance leakage parameter and the load impact coefficient according to a recursive updating rule based on the energy load observation record to obtain a predicted parameter vector of the current sampling period, and generating a self-learning safety margin value according to the predicted parameter vector and the temperature measurement value; Extrapolation is carried out on the lower boundary of the capacitor terminal voltage in the next prediction window according to the prediction parameter vector to obtain an available energy prediction value, an operation mark to be executed is generated according to an operation triggering event, and a task energy budget value is generated according to the mapping relation of the operation mark to be executed in a preset energy budget table; And when the voltage measured value of the capacitor terminal is larger than or equal to a preset wake-up voltage threshold value and passes the consistency check of the predicted parameter vector, the low-power-consumption sleep state is exited and the execution sequence corresponding to the operation identifier to be executed is restored.
2. The adaptive low power sleep control method for secondary fused capacitor powering of claim 1, wherein generating the energy load observation record comprises: when each preset sampling period starts, counting the preset sampling period by a local timer and increasing the sampling period index to obtain a current sampling period index; Triggering an analog-to-digital conversion channel and a temperature acquisition channel to sample a capacitor terminal voltage measured value and a temperature measured value in a sampling window corresponding to the sampling period index, and writing the capacitor terminal voltage measured value and the temperature measured value into a voltage field and a temperature field corresponding to the sampling period index; And packaging the sampling period index, the capacitor end voltage measured value, the temperature measured value and the load event count value into the energy load observation record according to a preset field sequence.
3. The adaptive low power consumption sleep control method for secondary capacitor charging according to claim 1 or 2, wherein the obtaining the prediction parameter vector of the current sampling period comprises: Reading capacitance end voltage measurement values and load event count values corresponding to adjacent two sampling period indexes from the energy load observation record, calculating capacitance end voltage difference components and load event difference components, and writing the capacitance end voltage difference components and the load event difference components into an update input item of a current sampling period; Based on the capacitance end voltage difference component and the forgetting factor, recursively correcting the capacitance equivalent leakage parameter of the previous sampling period to obtain the capacitance equivalent leakage parameter of the current sampling period; And carrying out recursive correction on the load impact coefficient of the previous sampling period based on the load event difference component, the capacitor end voltage difference component and the forgetting factor to obtain the load impact coefficient of the current sampling period, and splicing the capacitor equivalent leakage parameter and the load impact coefficient to form the prediction parameter vector.
4. The adaptive low power sleep control method for secondary fused capacitor powering of claim 3, wherein generating the self-learning safety margin value comprises: Extrapolation is carried out on the capacitor terminal voltage predicted value of the current sampling period based on the predicted parameter vector of the previous sampling period, and the capacitor terminal voltage predicted value is differenced with the capacitor terminal voltage measured value in the energy load observation record to obtain a voltage prediction error value; Generating a temperature weight based on a mapping result of the temperature measured value in a preset temperature interval table, writing the voltage prediction error value into an error buffer area according to the temperature weight, and carrying out exponential weighting summarization on a plurality of latest voltage prediction error values to obtain an error envelope value; And carrying out component weighting on the error envelope value according to the capacitance equivalent leakage parameter and the load impact coefficient in the prediction parameter vector to generate the self-learning safety margin value.
5. The adaptive low power sleep control method for secondary fused capacitor powering of claim 3, wherein obtaining the available energy prediction value comprises: Reading the prediction parameter vector and the preset sampling period, generating the window step number of the next prediction window, and writing the window step number into an extrapolation configuration item of the current sampling period; Generating a predicted load event count value based on a load event count value of a plurality of last sampling periods, taking a capacitor end voltage measured value in the energy load observation record as an initial voltage, accumulating voltage attenuation in the window step number according to the capacitor equivalent leakage parameter, generating a window impact deduction according to the load impact coefficient and the predicted load event count value, and superposing the voltage attenuation and the window impact deduction to obtain a capacitor end voltage lower boundary in a next predicted window; comparing the lower boundary of the capacitor terminal voltage with a preset minimum working voltage threshold, taking the larger of the lower boundary of the capacitor terminal voltage and the preset minimum working voltage threshold as a predicted terminal voltage, and performing energy conversion based on the predicted terminal voltage and the capacitor terminal voltage measured value to obtain the available energy predicted value.
6. The adaptive low power sleep control method for secondary fused capacitor powering of claim 1, wherein generating the task energy budget comprises: Reading the operation identifier to be executed, and searching a budget reference value and a budget correction rule identifier which are matched with the operation identifier to be executed in a preset energy budget table; According to the temperature measured value and the load event count value in the energy load observation record, determining a temperature correction coefficient and a load correction coefficient according to the budget correction rule mark; and respectively carrying out combination operation on the budget reference value, the temperature correction coefficient and the load correction coefficient to generate the task energy budget value.
7. The adaptive low power consumption sleep control method for secondary capacitor charging according to claim 1, wherein the low power consumption sleep state switching method comprises: Summing the task energy budget value and the self-learning safety margin value to obtain a gating threshold value, and comparing the gating threshold value with the available energy predicted value to generate a dormancy trigger mark; When the dormancy trigger mark indicates to be established, writing a dormancy state mark and the operation mark to be executed into a nonvolatile storage area, and writing a preset wake-up voltage threshold value into a wake-up configuration register; and after the writing of the sleep state identification is completed, sequentially closing the communication peripheral clock and the sampling peripheral clock, and switching the power supply state of the processor into a low-power-consumption sleep state.
8. The adaptive low power consumption sleep control method for secondary fused capacitor power taking according to claim 1, wherein the exiting the low power consumption sleep state and resuming the execution sequence corresponding to the operation identifier to be executed comprise: Periodically starting a sampling peripheral clock and collecting a capacitor terminal voltage measured value in a low-power-consumption sleep state, and comparing the capacitor terminal voltage measured value with a preset wake-up voltage threshold value to generate a wake-up candidate mark; When the wake-up candidate mark indicates to be established, a voltage prediction interval of a current sampling period is extrapolated according to the prediction parameter vector and the preset sampling period, and when the voltage measured value of the capacitor terminal falls into the voltage prediction interval, a consistency check mark is generated; When the consistency check mark indicates to be established, sequentially starting a processor main clock and a communication peripheral clock, and switching a power supply state from a low-power-consumption sleep state to an operation state; And after the power state switching is finished, reading the sleep state identification and the operation identification to be executed from the nonvolatile storage area, and recovering the corresponding execution sequence starting point according to the operation identification to be executed.
9. A self-adaptive low-power consumption dormancy control system for taking electricity by a secondary fusion capacitor is characterized by comprising the following components: One or more processors; A memory storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising the flow of the adaptive low power sleep control method for secondary fused capacitor power harvesting of any one of claims 1-8.
10. A computer readable medium storing software comprising instructions executable by one or more computers, the instructions causing the one or more computers to perform operations comprising the flow of the adaptive low power sleep control method for secondary fused capacitor power harvesting of any one of claims 1-8.

Description

Self-adaptive low-power-consumption dormancy control method for secondary fusion capacitor power taking Technical Field The invention relates to the technical field of power supply, in particular to a self-adaptive low-power-consumption dormancy control method for secondary fusion capacitor power supply. Background With the development of the edge sensing and distributed metering terminals towards the miniaturization and maintenance-free directions, micro-energy collection and secondary fusion power taking gradually become a common energy supply mode of a weak power supply scene, the terminals generally take a capacitor energy storage unit as energy buffering, the capacitor end voltage is influenced by fluctuation of power taking power, load pulse impact, equivalent leakage of devices and temperature drift to form non-stationary change, the existing low-power-consumption dormancy control is used for dormancy/awakening judgment according to a fixed threshold value or hysteresis threshold value of the capacitor end voltage, the structure is simple, but threshold value misjudgment is easy to occur when load event suddenly, temperature cross-region changes or power taking state suddenly changes, on one hand, false awakening occurs to cause secondary power failure and undervoltage reset to cause interruption of sampling and communication tasks, on the other hand, threshold value is excessively conserved for avoiding avoidance, the effective work duty ratio is compressed, and average standby duration is reduced. CN117120810a discloses a device for recording and storing measurement data related to an object or living body, namely a capacitor-powered measurement and storage device, which measures capacitor voltage, adjusts sampling frequency, storage frequency and transmission frequency based on measured voltage or predicted voltage, so as to maintain data recording and transmission under the condition of energy limitation, but its scheme lacks an online update mechanism for equivalent leakage change caused by temperature drift, and does not perform parameterized modeling and recursive update for voltage impact caused by load pulse, so that it is difficult to give a robust estimate for voltage lower bound in a prediction window, and it is difficult to introduce a self-learning safety margin based on prediction error to inhibit under-voltage reset risk, and erroneous judgment and insufficient stability may still occur when working condition is abrupt. CN108521837a discloses a power management method, a device and a micro energy power supply based on micro energy collection, the scheme is based on a static threshold value, lacks a technical structure for coupling capacitor terminal voltage change, load event impact and leakage temperature drift to the same recursive updating frame, does not give constraint of forming self-learning safety margin based on prediction error envelope, lacks task level budget expression for energy requirement of specific operation to be executed, and is difficult to map an operation trigger event and an energy budget table into task energy budget value and incorporate uniform gating, so that energy distribution fineness under complex tasks and emergency is insufficient. In summary, the existing dormancy control technology for capacitor power supply or micro-energy collection generally has the problems that threshold gating is sensitive to temperature drift, leakage change and load sudden impact, error awakening and undervoltage reset are easy to generate, interpretable prediction parameters are lacked to recursively update, voltage lower bounds in a prediction window are difficult to stably extrapolate, task-level energy budget constraint is lacked, average standby time is difficult to improve while awakening reliability is guaranteed, and in view of the problems existing in the prior art, the invention provides a self-adaptive low-power-consumption dormancy control method for secondary capacitor power taking, which is used for integrating prediction gating/awakening criteria of the capacitor power taking, namely, energy load observation records are generated through collecting capacitor end voltage measurement values, temperature measurement values and load event count values, a capacitor equivalent leakage parameter and a load impact coefficient are recursively updated to obtain a prediction parameter vector, a self-learning safety margin value is generated according to the prediction parameter vector, the prediction window voltage lower bounds are further extrapolated to obtain a usable energy prediction value, an operation triggering event is mapped to an operation mark to be executed to obtain a task energy budget value, and therefore the error-learning safety margin value is obtained, and the error-proofing scene under the conditions of the error-learning safety margin value is improved, and the error-proofing condition is relieved. Disclosure of Invention This section is intended to