Terminal energy consumption strategy adjustment method based on residual power prediction

Abstract

This invention relates to the field of energy management technology for IoT terminals and edge computing devices, specifically a terminal energy consumption strategy adjustment method based on remaining power prediction. It includes modules for data acquisition, phase space construction, state assessment, and control execution. The method acquires electrical signals, temperature, and energy replenishment prediction data to construct and update a virtual energy phase space. Its core is to calculate the frictional damping parameter characterizing capacity shrinkage based on ambient temperature to compress the reachable boundary of the space, and to calculate the energy manifold residual based on the actual and optimal discharge trajectories. When the residual exceeds a preset safety threshold, the method automatically reduces the sampling frequency and shuts down the peripheral interface. This invention achieves dynamic quantification of battery capacity decay limits in extreme environments, effectively overcoming dynamic mismatch, suppressing the risk of threshold breakdown caused by overestimation of power, and improving prediction accuracy.

Description

Technical Field

[0001] This invention relates to the field of energy management technology for Internet of Things (IoT) terminals and edge computing devices, specifically to a terminal energy consumption strategy adjustment method based on remaining power prediction. Background Technology

[0002] In the current smart grid and IoT operation and maintenance environment, edge computing nodes deployed in remote, cold and other complex working conditions often face irregular discharge scenarios such as sudden temperature drops and transient high current discharges; and due to the limitations of hardware resource configuration, such terminal devices usually lack dedicated power management chips.

[0003] To manage and predict energy consumption of terminal devices, existing solutions generally employ fixed capacity lookup tables or construct complex dynamic differential equations. While these solutions possess some evaluation capability under normal stable discharge scenarios, their high dependence on static empirical parameters and excessive computational consumption make them prone to dynamic mismatch and energy manifold topology mismatch when facing battery aging and complex climate changes. These defects lead to the risk of critical threshold breakdown in cold operating conditions due to overestimation of remaining power. At the same time, low-power microcontroller units struggle to handle the computational load of complex equations, easily causing resource conflicts between the computational consumption of the prediction algorithm and the remaining power.

[0004] Therefore, how to improve the accuracy of remaining power prediction and the reliability of adaptive energy consumption adjustment of terminal devices in complex operating conditions, under conditions where there are no dedicated power management chips and computing power is limited, has become an urgent technical problem to be solved. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a terminal energy consumption strategy adjustment method based on remaining power prediction. Specifically, the technical solution of this invention includes:

[0006] Acquire real-time electrical signal data, ambient temperature data, and environmental energy supply prediction data of the target device;

[0007] Based on preset initial potential energy parameters and preset energy decay model, a virtual energy phase space containing a state coordinate system is constructed, wherein the energy decay model characterizes the capacity decay boundary of the power supply module of the target device.

[0008] Based on the ambient temperature data, a frictional damping parameter characterizing the shrinkage of available capacity is calculated to compress the reachable boundary of the virtual energy phase space;

[0009] The current discharge power is calculated based on the real-time electrical signal data to update the current position coordinates of the target device in the virtual energy phase space;

[0010] Based on the environmental energy supply prediction data, the coordinates of the corresponding potential energy replenishment point are determined in the virtual energy phase space.

[0011] Calculate the optimal trajectory from the current position coordinates to the potential energy replenishment point coordinates;

[0012] Based on the actual discharge trajectory of the target device and the optimal state trajectory, the energy manifold residual characterizing the trajectory deviation is calculated;

[0013] If the energy manifold residual is greater than a preset safety threshold, a first control command is generated to reduce the sampling frequency of the analog-to-digital converter in the target device and shut down a preset peripheral interface.

[0014] If the energy manifold residual is less than or equal to the safety threshold, a second control command is generated to maintain the current sampling frequency of the analog-to-digital converter and the current power supply state of the peripheral interface.

[0015] Furthermore, the construction of a virtual energy phase space including a state coordinate system based on preset initial potential energy parameters and a preset energy decay model includes:

[0016] Configure the initial potential energy parameters as the initial state coordinates of the virtual energy phase space;

[0017] Extract a first polarization factor characterizing the degree of battery polarization and a second aging factor characterizing the degree of battery aging from the energy decay model;

[0018] The first polarization factor and the second aging factor are mapped to virtual elastic coefficients of the constraint state trajectory in the virtual energy phase space;

[0019] The virtual energy phase space is generated based on the initial state coordinates and the virtual elastic coefficients.

[0020] Further, the calculation of the frictional damping parameter characterizing the available capacity shrinkage based on the ambient temperature data to compress the reachable boundary of the virtual energy phase space includes:

[0021] Based on the preset temperature and capacity decay mapping relationship, the ambient temperature data is converted into the friction damping parameters;

[0022] Substitute the friction damping parameters into the boundary constraints of the virtual energy phase space;

[0023] Based on the substituted friction damping parameters, calculate the upper limit of the available capacity of the target device;

[0024] The available capacity upper limit is defined as the reachable boundary of the virtual energy phase space.

[0025] Further, the real-time electrical signal data includes real-time voltage data and real-time current data, and the step of calculating the current discharge power based on the real-time electrical signal data to update the current position coordinates of the target device in the virtual energy phase space includes:

[0026] The current discharge power is calculated based on the product of the real-time voltage data and the real-time current data;

[0027] Map the current discharge power to the current kinetic energy parameters of the target device;

[0028] Based on the current kinetic energy parameters, the current position coordinates of the target device in the virtual energy phase space are calculated.

[0029] Further, calculating the optimal state trajectory from the current position coordinates to the potential energy replenishment point coordinates includes:

[0030] Map the coordinates of the potential energy replenishment point to the target potential energy coordinates on the virtual energy phase space time axis;

[0031] Based on the principle of least action, an analytical geometric projection is performed on the energy topological manifold of the virtual energy phase space to solve for the extreme value of the difference between kinetic energy and potential energy on the time integral.

[0032] Based on the projection results, a Hamiltonian trajectory from the current position coordinates to the target potential energy coordinates is generated.

[0033] The Hamiltonian trajectory is determined as the optimal state trajectory.

[0034] Further, the calculation of the energy manifold residual characterizing the trajectory deviation based on the actual discharge trajectory of the target device and the optimal state trajectory includes:

[0035] Extract the historical location coordinate sequence of the target device within a preset historical time window;

[0036] The actual discharge trajectory is generated by fitting the historical location coordinate sequence.

[0037] Calculate the coordinate deviation between the actual discharge trajectory and the optimal state trajectory at a preset time node;

[0038] The coordinate deviation value is determined as the energy manifold residual.

[0039] Furthermore, the method also includes:

[0040] After executing the first control instruction or the second control instruction, the updated real-time electrical signal data of the target device is reacquired.

[0041] Based on the updated real-time electrical signal data, the updated discharge power is recalculated.

[0042] Based on the updated discharge power, the current position coordinates in the virtual energy phase space are iteratively updated.

[0043] Furthermore, the method also includes:

[0044] After executing the first control instruction to reduce the sampling frequency of the analog-to-digital converter, low-frequency stable discharge data is extracted;

[0045] The model parameters of the energy decay model are updated using the low-frequency stable discharge data, and the battery health assessment results are output.

[0046] Furthermore, the target device includes an IoT node device without a dedicated power management chip;

[0047] The environmental energy supply prediction data includes the probability data of cloudy and rainy weather within a future preset time period;

[0048] The preset peripheral interface includes a camera pin interface.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] 1. This method calculates the frictional damping parameter characterizing the shrinkage of available capacity based on ambient temperature data, and uses this parameter to compress the reachable boundary of the virtual energy phase space. This allows for the dynamic and objective quantification of the physical limitations on the capacity decay of battery power modules caused by extreme environments such as low temperatures. Compared with the traditional fixed capacity lookup table method, this mechanism effectively overcomes the dynamic mismatch and energy manifold topology mismatch that are easily generated during complex climate changes and battery aging. It fundamentally suppresses the risk of critical threshold breakdown caused by overestimating the remaining capacity under cold operating conditions, and improves the accuracy of capacity prediction.

[0051] 2. This method uses the principle of least action to perform analytical geometric projection on the energy topological manifold of the virtual energy phase space to solve for the extreme value of the difference between kinetic energy and potential energy on the time integral, generating a Hamiltonian trajectory as the optimal state trajectory. This approach abandons the traditional approach of constructing and solving complex dynamic differential equations, transforming complex dynamic prediction into a structured analytical solution on the phase space geometric manifold. This significantly reduces the computational load of the prediction algorithm and effectively overcomes the resource constraint bottleneck between prediction computing power consumption and remaining power in low-power devices without dedicated power management chips.

[0052] 3. This method calculates the energy manifold residual, which characterizes the trajectory deviation, by comparing the actual discharge trajectory with the optimal state trajectory. When the residual exceeds a preset safety threshold, it automatically generates control commands to reduce the sampling frequency of the analog-to-digital converter and shut down preset peripheral interfaces. This mechanism can accurately sense the severity of the system's deviation from the safe energy boundary caused by transient high-current discharge or irregular high-frequency tasks, and realizes adaptive triggering of bottom-level cascaded power consumption reduction. This not only ensures smooth power consumption reduction of the equipment during the power-off cycle, but also greatly improves the long-term survivability and online reliability of the terminal equipment in complex environments. Attached Figure Description

[0053] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0054] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0056] Example 1:

[0057] Please see Figure 1 A terminal energy consumption strategy adjustment method based on remaining power prediction includes:

[0058] Acquire real-time electrical signal data, ambient temperature data, and environmental energy supply prediction data of the target device;

[0059] Based on the preset initial potential energy parameters and the preset energy decay model, a virtual energy phase space containing a state coordinate system is constructed, wherein the energy decay model characterizes the capacity decay boundary of the power supply module of the target device.

[0060] Frictional damping parameters characterizing the shrinkage of available capacity are calculated based on ambient temperature data to compress the reachable boundary of the virtual energy phase space.

[0061] The current discharge power is calculated based on real-time electrical signal data to update the current position coordinates of the target device in the virtual energy phase space;

[0062] Based on environmental energy supply prediction data, the coordinates of the corresponding potential energy replenishment point are determined in the virtual energy phase space;

[0063] Calculate the optimal trajectory from the current position coordinates to the potential energy replenishment point coordinates;

[0064] Based on the actual discharge trajectory and the optimal state trajectory of the target device, the energy manifold residual characterizing the trajectory deviation is calculated.

[0065] If the energy manifold residual is greater than the preset safety threshold, a first control command is generated to reduce the sampling frequency of the analog-to-digital converter in the target device and shut down the preset peripheral interface.

[0066] If the energy manifold residual is less than or equal to the safety threshold, a second control command is generated to maintain the current sampling frequency of the analog-to-digital converter and the current power supply status of the peripheral interface.

[0067] This embodiment provides a terminal energy consumption strategy adjustment method based on remaining power prediction, which aims to solve the energy manifold topology mismatch problem of smart grid edge computing nodes deployed in remote mountainous areas in an environment without dedicated power management chips; the terminal energy consumption strategy adjustment method based on remaining power prediction acquires real-time electrical signal data, ambient temperature data and ambient energy replenishment prediction data of the target device to construct a data base for multi-dimensional state coordination;

[0068] Specifically, this method constructs a virtual energy phase space containing a state coordinate system based on preset initial potential energy parameters and preset energy decay model. The energy decay model characterizes the capacity decay boundary of the power supply module of the target device. This aims to transform the electrochemical kinetic mismatch phenomenon into a nonlinear phase space mapping, thereby establishing the absolute energy benchmark of the system.

[0069] This method calculates the frictional damping parameter, which characterizes the shrinkage of available capacity, based on ambient temperature data to compress the reachable boundary of the virtual energy phase space, ensuring that the objective constraints of the node thermodynamic state are quantified under cold operating conditions. During this period, the method calculates the current discharge power based on real-time electrical signal data to update the current position coordinates of the target device in the virtual energy phase space, characterizing instantaneous power fluctuations as kinetic energy transitions in the phase space. After the coordinate update is completed, the method determines the corresponding potential energy replenishment point coordinates in the virtual energy phase space based on ambient energy replenishment prediction data, calculates the optimal state trajectory from the current position coordinates to the potential energy replenishment point coordinates, and plans a smooth energy consumption reduction path that avoids the critical threshold of power depletion.

[0070] To achieve refined control, this method calculates the energy manifold residual, which characterizes the trajectory deviation, based on the actual discharge trajectory and the optimal state trajectory of the target device. When the energy manifold residual is greater than a preset safety threshold, this method generates a first control command to reduce the sampling frequency of the analog-to-digital converter in the target device and shut down the preset peripheral interface, thereby triggering the cascaded power saving mechanism of the underlying hardware.

[0071] The preset safety threshold setting logic is as follows: Before the target device leaves the factory, the minimum energy critical point for maintaining the normal operation of the core microcontroller unit and wake-up circuit of the target device is calibrated based on constant power discharge test; the potential energy coordinate value of the optimal state trajectory at the corresponding moment of the minimum energy critical point is multiplied by the preset fault tolerance margin coefficient, and the product result is calibrated as the safety threshold.

[0072] Conversely, when the energy manifold residual is less than or equal to the safety threshold, the method generates a second control command to maintain the current sampling frequency of the analog-to-digital converter and the current power supply state of the peripheral interface. This embodiment demonstrates the energy consumption adaptive adjustment capability under complex alternating climate conditions and verifies the adaptability of the technical solution under different discharge conditions.

[0073] Based on preset initial potential energy parameters and a preset energy decay model, a virtual energy phase space containing a state coordinate system is constructed, including:

[0074] Configure the initial potential energy parameters as the initial state coordinates of the virtual energy phase space;

[0075] The energy decay model extracts a first polarization factor characterizing the degree of battery polarization and a second aging factor characterizing the degree of battery aging. Specifically, the construction process of the preset energy decay model is as follows: Before the target device leaves the factory, a complete charge-discharge life cycle test is conducted on the same batch of power supply modules under different ambient temperature gradients, and the measured internal resistance data at different ambient temperatures and different charge-discharge cycle numbers are collected. Using ambient temperature and historical charge-discharge cycle numbers as independent variables and the collected measured internal resistance data as dependent variables, a multidimensional data mapping lookup table is constructed, and the multidimensional data mapping lookup table is stored in the non-volatile memory of the target device as the preset energy decay model. The aforementioned extraction of the internal resistance observation value at the current ambient temperature is to input the current ambient temperature and the currently recorded historical charge-discharge cycle number into the multidimensional data mapping lookup table to match and obtain the corresponding internal resistance value.

[0076] The first polarization factor and the second aging factor are mapped to virtual elastic coefficients of the constrained state trajectory in the virtual energy phase space;

[0077] A virtual energy phase space is generated based on the initial state coordinates and virtual elastic coefficients.

[0078] This embodiment is a further specification of the steps for constructing a virtual energy phase space containing a state coordinate system based on preset initial potential energy parameters and preset energy decay model; the terminal energy consumption strategy adjustment method based on remaining power prediction configures the initial potential energy parameters as the initial state coordinates of the virtual energy phase space to anchor the initial energy topology benchmark of the edge gateway node for high-frequency data reporting.

[0079] Based on this, the method deeply analyzes the complex solid electrolyte interface film thickening effect, extracting a first polarization factor characterizing the degree of battery polarization and a second aging factor characterizing the degree of battery aging from the energy decay model. Specifically, the system extracts the observed internal resistance value at the current ambient temperature from the energy decay model and divides it by the factory-preset reference internal resistance value to obtain a dimensionless polarization ratio as the first polarization factor; at the same time, it extracts the historical charge-discharge cycle count and divides it by the preset nominal design cycle life count to obtain a dimensionless aging ratio as the second aging factor; by introducing a reference value for division, the parameters that originally had different physical dimensions and numerical magnitudes are transformed into a unified dimensionless feature, eliminating the dimensional conflict in subsequent calculations.

[0080] To eliminate the estimation bias caused by long-term irreversible capacity loss, this method maps the first polarization factor and the second aging factor to the virtual elastic coefficients of the constrained state trajectory in the virtual energy phase space. This mapping process abandons the traditional fixed weight coefficients and adopts a weighted fusion logic based on the dynamic evolution of the service cycle, taking into account the irregular discharge characteristics of edge computing nodes.

[0081] Specifically, the system's preset basic polarization weight, reference temperature difference span, and basic aging weight are not empirical estimates, but rather fixed reference parameters determined by conducting multi-temperature gradient charge-discharge cycle calibration experiments covering the entire life cycle of the target equipment before it leaves the factory, and by using the nonlinear least squares method to fit and optimize the experimental data, with the goal of minimizing the mean square error between the predicted internal resistance and the measured internal resistance.

[0082] Based on this, the system positively correlates the weight of the first polarization factor with the average daily ambient temperature difference of the target device over the past preset natural days. The system calculates the temperature fluctuation ratio by dividing the average daily ambient temperature difference by the preset baseline temperature difference span, and multiplies this ratio by the basic polarization weight to obtain the dynamically adjusted weight of the first polarization factor. The polarization weight increases accordingly with the increase of the temperature fluctuation ratio to amplify the compression effect of the increase in short-term transient internal resistance on the phase space boundary. The weight of the second aging factor is nonlinearly and exponentially bound to the cumulative discharge depth of the target device since it leaves the factory. Specifically, the aging aggravation coefficient is calculated with the natural constant as the base and the cumulative discharge depth percentage as the exponent. This coefficient is multiplied by the basic aging weight to obtain the dynamic weight of the second aging factor. The aging weight increases nonlinearly and exponentially with the increase of the cumulative discharge depth, thereby dominating the permanent weakening of the phase space elasticity.

[0083] The formula for calculating the dynamic weight of the second aging factor is:

[0084]

[0085] in, The dynamic weights characterizing the second aging factor Characterized by the base aging weight, e is the natural constant. Characterized by the percentage of cumulative discharge depth;

[0086] The system multiplies the first polarization factor with the dynamically adjusted weight of the first polarization factor, multiplies the second aging factor with the dynamic weight of the second aging factor, and sums the products of the two. The sum is then used directly as the virtual elasticity coefficient. This design accurately distinguishes the different physical mechanisms of short-term polarization and long-term aging on battery capacity decay.

[0087] This method generates a virtual energy phase space based on initial state coordinates and virtual elasticity coefficients; specifically, it constructs a virtual energy phase space based on potential energy. Let be a two-dimensional coordinate system with the vertical axis and the equivalent discharge rate v as the horizontal axis, and let the virtual elastic coefficient k be mapped to the equivalent mass parameter characterizing the energy conversion inertia of the target device in phase space. The phase space boundary equation is defined as follows:

[0088]

[0089] in, The term in the formula represents the total boundary energy of the virtual energy phase space. This is represented by the current kinetic energy parameters of the target device. ; Calculate the physical constants for kinetic energy;

[0090] The introduction of the virtual elastic coefficient enables the phase space boundary contraction to respond in real time to the nonlinear decrease in the battery's ability to resist discharge shocks, making the subsequently calculated state trajectory smoother and effectively improving the dynamic mismatch problem caused by the traditional fixed capacity lookup table method after battery aging; the scheme shows good parameter fitting robustness in dealing with the nonlinear capacity decay scenario after long-term service of edge nodes.

[0091] Frictional damping parameters characterizing the shrinkage of available capacity are calculated based on ambient temperature data to compress the reachable boundary of the virtual energy phase space, including:

[0092] Based on the preset temperature and capacity decay mapping relationship, the ambient temperature data is converted into friction damping parameters;

[0093] Substitute the friction damping parameters into the boundary constraints of the virtual energy phase space;

[0094] Based on the substituted friction damping parameters, calculate the upper limit of the available capacity of the target equipment;

[0095] The upper limit of available capacity is defined as the reachable boundary of the virtual energy phase space.

[0096] This embodiment is a further specification of the step of calculating the friction damping parameter characterizing the shrinkage of available capacity based on ambient temperature data to compress the reachable boundary of the virtual energy phase space; for the sudden temperature drop conditions often encountered by hydrological monitoring stations in high-altitude and cold regions, this terminal energy consumption strategy adjustment method based on remaining power prediction is based on a preset temperature and capacity decay mapping relationship, and converts ambient temperature data into friction damping parameters, aiming to accurately quantify the physical limitation of low temperature environment on the reduction of electrochemical reaction activity inside the battery;

[0097] The aforementioned preset temperature-capacity decay mapping relationship is not a simple linear ratio, but a three-stage damping mapping logic constructed based on the physical characteristic of increased electrolyte viscosity at low temperatures in lithium-ion batteries. The system sets 0 degrees Celsius and -15 degrees Celsius as the first and second damping transition thresholds, respectively. When the ambient temperature is higher than or equal to 0 degrees Celsius, the system fixes the friction damping parameter to a minimum value to maintain full capacity. When the temperature is lower than 0 degrees Celsius and higher than or equal to -15 degrees Celsius, the friction damping parameter is mapped to the temperature drop rate in a positive first-order manner to characterize the uniform slowdown of ion diffusion rate.

[0098] Specifically, the system sets the basic temperature drop step value for this range to 1 degree Celsius, corresponding to a basic increment of 0.05 for the friction damping parameter. That is, for every 1 degree Celsius drop in ambient temperature from 0 degrees Celsius, the friction damping parameter increases by 0.05 from 0. For example, when the ambient temperature is -10 degrees Celsius, the friction damping parameter is precisely assigned a value of 0.50. When the temperature is below -15 degrees Celsius, due to the sharp decrease in the utilization rate of active materials, the system switches the rate of increase of the friction damping parameter to a second-order jump mode, rapidly increasing the friction damping parameter.

[0099] Specifically, the system uses the friction damping parameter of 0.75 at -15 degrees Celsius as the starting point for the jump. Afterward, for every 1 degree Celsius decrease in temperature, the increment of the friction damping parameter is no longer constant, but rather the product of the previous increment and a fixed second-order deterioration coefficient, such as 1.2. For example, the increment is 0.06 when decreasing from -15 degrees Celsius to -16 degrees Celsius, and 0.072 when decreasing to -17 degrees Celsius. This causes the friction damping parameter to increase rapidly with decreasing temperature, and stops increasing when it reaches its extreme value of 1.0, limiting it to 1.0 to prevent overflow in boundary calculations. After the transformation, this method substitutes the friction damping parameter into the boundary constraints of the virtual energy phase space, thereby constructing a cascade mechanism between thermodynamic friction damping and the upper limit of available potential energy.

[0100] This method, based on the substituted friction damping parameters, calculates the upper limit of the available capacity of the target equipment, objectively reflecting the transient contraction of available energy caused by low temperatures; specifically, it includes: setting the factory nominal capacity of the target equipment as... The friction damping parameters are The upper limit of available capacity under the current environment can be directly calculated using the boundary constraint formula. The boundary constraint formula is:

[0101]

[0102] This method defines the upper limit of available capacity as the reachable boundary of the virtual energy phase space. Through this series of nonlinear phase space mappings, the technical solution effectively suppresses the risk of critical threshold breakdown caused by overestimation of remaining power under cold operating conditions, and verifies the system's state convergence capability under extreme temperature difference environments.

[0103] Real-time electrical signal data includes real-time voltage data and real-time current data. The current discharge power is calculated based on this data to update the target device's current position coordinates in the virtual energy phase space, including:

[0104] The current discharge power is calculated based on the product of real-time voltage data and real-time current data.

[0105] Map the current discharge power to the current kinetic energy parameters of the target device;

[0106] Based on the current kinetic energy parameters, calculate the current position coordinates of the target device in the virtual energy phase space.

[0107] This embodiment further specifies the step of calculating the current discharge power based on real-time electrical signal data to update the current position coordinates of the target device in the virtual energy phase space. In the scenario of a security monitoring terminal performing a sudden image capture task, the real-time electrical signal data specifically includes real-time voltage data and real-time current data synchronously sampled by an analog-to-digital converter. The terminal energy consumption strategy adjustment method based on remaining power prediction calculates the current discharge power based on the product of real-time voltage data and real-time current data to capture the drastic fluctuations in system state caused by transient high-current discharge.

[0108] This method maps the current discharge power to the current kinetic energy parameters of the target device. The mapping logic converts the instantaneous power into energy consumption within a specific time integration step, which characterizes the rate at which the target device transitions to a lower energy state in phase space.

[0109] Specifically, the system directly maps the dimensionless normalized value of the current discharge power to the equivalent discharge rate v on the horizontal axis, where the dimensionless normalized value is the quotient obtained by dividing the current discharge power by the maximum rated discharge power preset by the target device, thereby eliminating the physical dimension of power.

[0110] And based on the equivalent mass parameter k, through the formula Calculate the current kinetic energy parameters To establish the position of the current coordinate point on the horizontal axis of the phase space;

[0111] Based on the current kinetic energy parameters, this method calculates the current position coordinates of the target device in the virtual energy phase space. The increase in kinetic energy parameters is equivalent to the rapid consumption of the system's potential energy. In order to calculate the vertical axis coordinate, the system synchronously multiplies the current discharge power by the sampling time interval of the analog-to-digital converter to calculate the absolute energy consumption value within the discrete period.

[0112] This method calculates the displacement of the coordinate point on the potential energy axis in phase space. Based on the scaling ratio of the virtual energy phase space, the system divides the absolute energy consumption value by the total available energy reference of the target device to obtain the percentage decrease in potential energy corresponding to the absolute energy consumption within the discrete period. The total available energy reference is the upper limit of the current available capacity calculated based on the friction damping parameters. The product of the absolute energy consumption value and the nominal operating voltage of the target equipment is used to ensure that the absolute energy consumption value has the same energy dimension as the energy reference.

[0113] Then, subtract the displacement corresponding to the percentage decrease in potential energy from the previous potential energy coordinate to complete the independent update of the vertical coordinate;

[0114] Furthermore, by multiplying the relative energy consumption percentage by the total length of the phase space potential energy axis, the single-step descent displacement of the coordinate point on the potential energy axis can be accurately calculated. The system subtracts this descent displacement from the position coordinates of the previous moment, thereby transforming the discrete electrical signal fluctuations into continuous phase space geometric coordinate changes in situ. Through this direct dimensionality reduction mapping from physical quantities to geometric quantities, the system does not need to construct complex dynamic differential equations, and can achieve real-time iteration of the phase space trajectory within a low-power microcontroller unit by relying only on basic arithmetic operations. This processing mechanism provides a unified mathematical evaluation basis for subsequent trajectory comparison, reflecting the balance between computing power and accuracy under pulse load interference.

[0115] Calculate the optimal trajectory from the current position coordinates to the potential energy replenishment point coordinates, including:

[0116] Map the coordinates of the potential energy replenishment point to the target potential energy coordinates on the virtual energy phase space time axis;

[0117] Based on the principle of least action, an analytical geometric projection is performed on the energy topological manifold of the virtual energy phase space to solve for the extreme value of the difference between kinetic energy and potential energy on the time integral.

[0118] Based on the projection results, a Hamiltonian trajectory from the current position coordinates to the target potential energy coordinates is generated;

[0119] The Hamiltonian trajectory is determined as the optimal state trajectory.

[0120] This embodiment is a further specification of the steps for calculating the optimal state trajectory from the current position coordinates to the potential energy replenishment point coordinates; facing the harsh working conditions of agricultural IoT sensor nodes that rely on solar power and encounter continuous rainy weather, this terminal energy consumption strategy adjustment method based on remaining power prediction maps the potential energy replenishment point coordinates to the target potential energy coordinates on the virtual energy phase space time axis, so as to establish the minimum safe power reserve boundary that the system must retain when the weather is expected to clear.

[0121] Under this constraint, the method is based on the principle of least action and performs analytical geometric projection on the energy topology manifold of the virtual energy phase space to solve for the extreme value of the difference between kinetic energy and potential energy on the time integral. Specifically, the system takes the current time corresponding to the current position coordinate as the evaluation starting point and the expected environmental energy replenishment time corresponding to the target potential energy coordinate as the evaluation ending point, thereby defining the complete time integral span.

[0122] Within this time span, the system defines the product of the current discharge power and the step time in each preset discrete time step as kinetic energy, which is used to characterize the severity of transient energy consumption. At the same time, the difference between the remaining available capacity of the target device at the corresponding step and the safe energy baseline at the target potential energy coordinate is defined as potential energy, which is used to characterize the energy reserve margin for the system to maintain normal operation. To achieve path comparison, the system arranges and combines multiple available sampling frequency gradients supported by the target device and the switching states of various peripheral interfaces to generate multiple candidate working modes that characterize different energy reduction efforts.

[0123] For each candidate operating mode, combined with the preset static bottom current of the equipment, the expected power time series of the mode within the above time span is calculated, and multiple alternative discharge paths that evolve towards the target potential energy coordinate in phase space are generated by integration.

[0124] The system calculates and accumulates the difference between kinetic energy and potential energy at each time step. By comparing the sum of all candidate discharge paths, the path with the smallest absolute value of the sum is selected as the extreme solution. This projection operation aims to find an evolution path that minimizes the total energy loss of the system and minimizes the discharge power fluctuation during the entire waiting period. This transforms the traditional matrix iterative filtering algorithm into a structured analytical solution on a geometric manifold, reducing the computational burden on the low-power microcontroller unit.

[0125] Based on the projection results, this method generates a Hamiltonian trajectory from the current position coordinates to the target potential energy coordinates. Specifically, it includes: taking the discrete coordinate points on the extreme value solution path with the smallest absolute value of the cumulative sum as control nodes, using a cubic spline interpolation algorithm to perform polynomial fitting on each control node, and using natural boundary conditions in the cubic spline interpolation algorithm to make the second derivative values ​​of the first and last two points of the fitted curve zero, generating a smooth curve that is continuous and differentiable everywhere, and determining the curve as the Hamiltonian trajectory.

[0126] The generation process does not call a complex differential equation solver, but directly connects the selected extreme solution paths in the virtual energy phase space. The principle is that in physics, the Hamiltonian trajectory represents the energy conservation path in a conservative system. In this business scenario, it is mapped to the ideal consumption baseline of the device approaching the safe power limit at the most stable discharge rate during the period without replenishment. The trajectory strictly avoids the high kinetic energy steep slope range that represents transient high current discharge in terms of geometric shape.

[0127] This method determines the Hamiltonian trajectory as the optimal state trajectory; this trajectory serves as a feedforward baseline, ensuring that the power consumption planning of the equipment conforms to the objective laws of external energy replenishment, and improving the probability of continuous operation of the system during periods without energy replenishment.

[0128] Based on the actual discharge trajectory and the optimal state trajectory of the target device, the energy manifold residual characterizing the trajectory deviation is calculated, including:

[0129] Extract the historical location coordinate sequence of the target device within a preset historical time window;

[0130] The actual discharge trajectory is generated by fitting historical location coordinate sequences.

[0131] Calculate the coordinate deviation between the actual discharge trajectory and the optimal state trajectory at a preset time node;

[0132] The coordinate deviation value is defined as the energy manifold residual.

[0133] This embodiment is a further specification of the step of calculating the energy manifold residual representing the trajectory deviation based on the actual discharge trajectory and the optimal state trajectory of the target device; in the scenario where the energy consumption is aggravated due to the intensive environmental data sampling task performed by the meteorological buoy, the terminal energy consumption strategy adjustment method based on the prediction of remaining power extracts the historical position coordinate sequence of the target device within a preset historical time window to obtain a discrete phase space coordinate dataset reflecting the recent discharge trend.

[0134] This method generates the actual discharge trajectory based on the historical position coordinate sequence. It eliminates the local coordinate jump caused by single sampling noise by using the least squares method or polynomial smoothing algorithm to ensure the smoothness of the trajectory representation. The technical motivation for using the polynomial smoothing algorithm here is that the sudden high-frequency sampling of the edge computing node will cause violent sawtooth oscillations in the coordinate sequence. Directly comparing with the original discrete sequence will lead to false triggering of power reduction commands. Polynomial smoothing can effectively filter out transient high-frequency fluctuations and extract the main trend of energy decay.

[0135] Based on this, the method calculates the coordinate deviation between the actual discharge trajectory and the optimal state trajectory at a preset time node. Specifically, the system does not only calculate the absolute deviation at a single current moment, but extracts the ordinate values ​​of the actual discharge trajectory and the optimal state trajectory at multiple recent consecutive discrete time nodes, and calculates the absolute value of the ordinate difference between each pair of corresponding nodes.

[0136] The system assigns a specific decay weight to each discrete time node based on the time distance. Specifically, the system adopts an exponential decay allocation logic, with the current time as the zero distance reference point. For each fixed time step increase in the time difference between the historical time node and the present, the corresponding weight is multiplied by a fixed decay coefficient between 0 and 1 based on the weight of the previous closer node. For example, 0.8 is selected as the smooth decay reference.

[0137] Meanwhile, to ensure the mathematical rigor of deviation quantification and avoid scale divergence, the system normalizes the basic weights assigned to all extraction nodes. That is, the basic weight of each node is divided by the sum of the basic weights of all nodes, and the sum of the actual weights participating in the final weighting is forced to be always equal to 1. For example, when extracting the three most recent time nodes, if the basic weights are 1, 0.8 and 0.64 respectively, the sum is 2.44, and the normalized actual weights are 0.41, 0.33 and 0.26 respectively. This weight allocation logic intuitively reflects that the more recent the trajectory deviation, the greater its impact on the assessment of the system's current energy consumption status.

[0138] The quantified coordinate deviation value can be obtained by multiplying the absolute value of the difference between the ordinates of each pair of corresponding nodes by the actual weight of that node, and summing the weighted absolute values.

[0139] The formula for calculating the quantified coordinate deviation value is as follows:

[0140]

[0141] in, The coordinate deviation value representing quantization, i.e., the energy manifold residual, and n represents the total number of discrete time nodes extracted. The normalized actual weights at the i-th time point are represented, where i is a positive integer; It is an absolute value symbol, and its range of values ​​is... ; The ordinate value representing the actual discharge trajectory at the i-th time node. The ordinate value representing the optimal state trajectory at the i-th time node;

[0142] This calculation step quantifies the severity of the target device's actual remaining power falling short of the theoretical safe power due to the recent execution of too many high-energy-consuming tasks;

[0143] This method determines the coordinate deviation value as the energy manifold residual, providing an accurate numerical basis for triggering multi-level energy reduction commands. This multi-node weighted deviation calculation method not only captures the static spatial distance between the current power level and the safety baseline, but also incorporates the historical cumulative trend of power deviation. This residual evaluation mechanism objectively reflects the degree to which the system energy state deviates from the expected safety boundary, greatly improving the reliability and sensitivity of the system's state perception and early warning under dynamic load conditions.

[0144] Example 2:

[0145] The method also includes:

[0146] After executing the first control command or the second control command, the updated real-time electrical signal data of the target device is reacquired.

[0147] Based on the updated real-time electrical signal data, the updated discharge power is recalculated.

[0148] Based on the updated discharge power, the current position coordinates in the virtual energy phase space are iteratively updated.

[0149] This embodiment provides a closed-loop feedback regulation mechanism. In a scenario where a geological disaster early warning node in a remote area cuts off the power supply to a high-power communication module due to a power consumption reduction strategy, the terminal power consumption strategy regulation method based on remaining power prediction reacquires updated real-time electrical signal data of the target device after executing the first control command or the second control command, in order to obtain the transient response after underlying hardware physical intervention or state maintenance. When the first control command is executed, the current data collected by the system decreases significantly due to the triggering of frequency reduction and peripheral interface shutdown. When the second control command is executed, the current data collected by the system maintains a stable baseline state.

[0150] This method recalculates the updated discharge power based on the updated real-time electrical signal data, accurately reflecting the energy consumption reduction effect brought about by the frequency reduction and power-off operation or the energy consumption continuation under the state maintenance.

[0151] This method iteratively updates the current position coordinates in the virtual energy phase space based on the updated discharge power. After executing the first control command, as the discharge power representing kinetic energy decreases significantly, the rate at which the system slides towards lower potential energy in the phase space is effectively curbed, causing the slope of the actual discharge trajectory to gradually decrease and approach the pre-planned optimal state trajectory. After executing the second control command, the trajectory continues to extend at a steady slope. This iterative update mechanism realizes the adaptive adjustment and state convergence closed loop of the system under harsh operating conditions, verifying the effectiveness of the closed-loop control strategy in suppressing accelerated power loss.

[0152] Example 3:

[0153] The method also includes:

[0154] After executing the first control command to reduce the sampling frequency of the analog-to-digital converter, low-frequency stable discharge data is extracted;

[0155] The model parameters of the energy decay model are updated using low-frequency stable discharge data, and the battery health assessment results are output.

[0156] This embodiment provides a reverse model calibration mechanism based on changes in the physical environment of control actions. In a scenario where a wildlife tracking collar enters a dormant monitoring mode, the terminal energy consumption strategy adjustment method based on remaining power prediction extracts low-frequency stable discharge data after executing a first control command to reduce the sampling frequency of the analog-to-digital converter. Since the first control command shuts down power-consuming peripherals, the large current pulse interference inside the system is physically eliminated. The extracted low-frequency stable discharge data excludes the rebound noise interference of high-frequency polarization voltage and is closer to the actual open-circuit voltage characteristics of the battery. This method uses the low-frequency stable discharge data to update the model parameters of the energy decay model and uses the pure discharge data as a high-confidence sample input parameter identification module to adaptively correct the aging factors that characterize long-term life decay.

[0157] Specifically, the system iteratively extracts the average voltage drop of the low-frequency stable discharge data over multiple consecutive independent preset time periods according to a preset sampling cycle, in order to construct a voltage drop sequence containing multiple independent time-series samples. Each voltage drop is then compared with the factory-preset reference static voltage drop. If the extracted voltage drop sequence shows three consecutive values ​​exceeding the reference static voltage drop, and each deviation exceeds a preset aging threshold, the system determines that the self-discharge rate caused by an internal micro-short circuit or material phase transition has substantially degraded. The aging threshold is 15% of the factory-preset reference static voltage drop.

[0158] At this point, the system triggers the model parameter update mechanism, divides the average of the three consecutive deviations by the reference static voltage drop, calculates the dimensionless voltage degradation ratio, multiplies the voltage degradation ratio by the preset degradation conversion coefficient to obtain the dimensionless compensation gain, and adds the compensation gain to the second aging factor in the energy decay model, so that the model enters the accelerated decay assessment stage in advance.

[0159] This processing method not only eliminates the dimensional conflict between the voltage scaling dimension and the dimensionless aging factor, but also objectively quantifies the equivalent reduction in overall lifespan due to self-discharge degradation through proportional mapping. After the correction is completed, the system calculates the current available capacity limit based on the updated second aging factor and the current ambient temperature.

[0160] Specifically, the system uses the factory-nominated total capacity as a baseline value. It then uses the current ambient temperature to look up a preset temperature-capacity retention rate table to obtain the upper limit of the physically usable capacity at the current temperature. The system introduces an updated second aging factor as a capacity loss penalty. The calculation logic for this penalty is as follows: subtract a preset health baseline value (e.g., 1.0) from the second aging factor, and multiply this difference by a preset attenuation sensitivity weight (e.g., 0.15) to obtain the corresponding capacity loss ratio. Subtracting the capacity value corresponding to this capacity loss ratio from the aforementioned upper limit of physically usable capacity yields the current upper limit of usable capacity. For example, if the upper limit of physically usable capacity at the current temperature is 90% of the nominal total capacity, and the updated second aging factor is 1.2, then the capacity loss ratio is 0.03 (0.2 multiplied by 0.15), and the final current upper limit of usable capacity is 87% of the nominal total capacity.

[0161] The system calculates the quotient between the current available capacity limit and the factory-nominated total capacity, and defines this quotient as the model's remaining capacity ratio;

[0162] This method outputs battery health assessment results based on the updated model's remaining capacity ratio. For example, when the ratio is below 80%, it outputs an alarm prompting that the battery needs to be replaced. This scheme avoids the pollution of model parameters by high-frequency discharge polarization artifacts through the in-situ construction of a low-noise observation environment with clear judgment criteria and cumulative update rules. It reversely improves the observation accuracy of the algorithm model and realizes a complete closed loop from the evolution of the underlying algorithm to the top-level business interaction, demonstrating the ability to automatically optimize parameters without human intervention.

[0163] Example 4:

[0164] The target devices include IoT node devices without dedicated power management chips;

[0165] Environmental energy supply forecast data includes weather and rain probability data for a future preset period;

[0166] The default peripheral interfaces include a camera pin interface.

[0167] This embodiment is a refined adaptation to specific physical entities and business scenarios. In this solution, the target devices include IoT node devices without dedicated power management chips, such as forest fire prevention image monitoring nodes deployed in deep mountain forest areas and powered only by micro solar panels. Although this hardware architecture simplifies the system hardware overhead, it is very easy to cause resource conflicts between the computing power consumption of the prediction algorithm and the remaining power.

[0168] To address this issue, the terminal energy consumption strategy adjustment method based on remaining power prediction obtains environmental energy replenishment prediction data, including the probability of cloudy / rainy weather within a preset future time period. This data serves as a feedforward constraint, directly determining the time span for the system to set potential energy replenishment points in the virtual energy phase space. When facing abnormal operating conditions where the energy manifold residual exceeds the limit, the method controls preset peripheral interfaces, including camera pin interfaces, by directly pulling down the level of general-purpose input / output pins to cut off the physical power supply to the camera. This technical solution deeply integrates top-level weather prediction, bottom-level phase space algorithms, and specific hardware pin control. Under limited hardware conditions, it improves the online rate of IoT nodes in long-term severe weather environments, verifying the practical engineering significance of the software-hardware collaborative energy reduction mechanism.

[0169] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A terminal energy consumption strategy adjustment method based on remaining power prediction, characterized in that, The method includes: Acquire real-time electrical signal data, ambient temperature data, and environmental energy supply prediction data of the target device; Based on preset initial potential energy parameters and preset energy decay model, a virtual energy phase space containing a state coordinate system is constructed, wherein the energy decay model characterizes the capacity decay boundary of the power supply module of the target device. Based on the ambient temperature data, a frictional damping parameter characterizing the shrinkage of available capacity is calculated to compress the reachable boundary of the virtual energy phase space; The current discharge power is calculated based on the real-time electrical signal data to update the current position coordinates of the target device in the virtual energy phase space; Based on the environmental energy supply prediction data, the coordinates of the corresponding potential energy replenishment point are determined in the virtual energy phase space. Calculate the optimal trajectory from the current position coordinates to the potential energy replenishment point coordinates; Based on the actual discharge trajectory of the target device and the optimal state trajectory, the energy manifold residual characterizing the trajectory deviation is calculated; If the energy manifold residual is greater than a preset safety threshold, a first control command is generated to reduce the sampling frequency of the analog-to-digital converter in the target device and shut down a preset peripheral interface. If the energy manifold residual is less than or equal to the safety threshold, a second control command is generated to maintain the current sampling frequency of the analog-to-digital converter and the current power supply state of the peripheral interface.

2. The terminal energy consumption strategy adjustment method based on remaining power prediction according to claim 1, characterized in that, The virtual energy phase space, which includes a state coordinate system, is constructed based on preset initial potential energy parameters and a preset energy decay model, including: Configure the initial potential energy parameters as the initial state coordinates of the virtual energy phase space; Extract a first polarization factor characterizing the degree of battery polarization and a second aging factor characterizing the degree of battery aging from the energy decay model; The first polarization factor and the second aging factor are mapped to virtual elastic coefficients of the constraint state trajectory in the virtual energy phase space; The virtual energy phase space is generated based on the initial state coordinates and the virtual elastic coefficients.

3. The terminal energy consumption strategy adjustment method based on remaining power prediction according to claim 1, characterized in that, The calculation of frictional damping parameters characterizing the available capacity shrinkage based on the ambient temperature data, in order to compress the reachable boundary of the virtual energy phase space, includes: Based on the preset temperature and capacity decay mapping relationship, the ambient temperature data is converted into the friction damping parameters; Substitute the friction damping parameters into the boundary constraints of the virtual energy phase space; Based on the substituted friction damping parameters, calculate the upper limit of the available capacity of the target device; The available capacity upper limit is defined as the reachable boundary of the virtual energy phase space.

4. The terminal energy consumption strategy adjustment method based on remaining power prediction according to claim 1, characterized in that, The real-time electrical signal data includes real-time voltage data and real-time current data. The step of calculating the current discharge power based on the real-time electrical signal data to update the current position coordinates of the target device in the virtual energy phase space includes: The current discharge power is calculated based on the product of the real-time voltage data and the real-time current data; Map the current discharge power to the current kinetic energy parameters of the target device; Based on the current kinetic energy parameters, the current position coordinates of the target device in the virtual energy phase space are calculated.

5. The terminal energy consumption strategy adjustment method based on remaining power prediction according to claim 1, characterized in that, The calculation of the optimal state trajectory from the current position coordinates to the potential energy replenishment point coordinates includes: Map the coordinates of the potential energy replenishment point to the target potential energy coordinates on the virtual energy phase space time axis; Based on the principle of least action, an analytical geometric projection is performed on the energy topological manifold of the virtual energy phase space to solve for the extreme value of the difference between kinetic energy and potential energy on the time integral. Based on the projection results, a Hamiltonian trajectory from the current position coordinates to the target potential energy coordinates is generated. The Hamiltonian trajectory is determined as the optimal state trajectory.

6. The terminal energy consumption strategy adjustment method based on remaining power prediction according to claim 1, characterized in that, The calculation of the energy manifold residual characterizing the trajectory deviation based on the actual discharge trajectory of the target device and the optimal state trajectory includes: Extract the historical location coordinate sequence of the target device within a preset historical time window; The actual discharge trajectory is generated by fitting the historical location coordinate sequence. Calculate the coordinate deviation between the actual discharge trajectory and the optimal state trajectory at a preset time node; The coordinate deviation value is determined as the energy manifold residual.

7. The terminal energy consumption strategy adjustment method based on remaining power prediction according to claim 1, characterized in that, The method further includes: After executing the first control instruction or the second control instruction, the updated real-time electrical signal data of the target device is reacquired. Based on the updated real-time electrical signal data, the updated discharge power is recalculated. Based on the updated discharge power, the current position coordinates in the virtual energy phase space are iteratively updated.

8. The terminal energy consumption strategy adjustment method based on remaining power prediction according to claim 1, characterized in that, The method further includes: After executing the first control instruction to reduce the sampling frequency of the analog-to-digital converter, low-frequency stable discharge data is extracted; The model parameters of the energy decay model are updated using the low-frequency stable discharge data, and the battery health assessment results are output.

9. The terminal energy consumption strategy adjustment method based on remaining power prediction according to claim 1, characterized in that, The target devices include IoT node devices without dedicated power management chips; The environmental energy supply prediction data includes the probability data of cloudy and rainy weather within a future preset time period; The preset peripheral interface includes a camera pin interface.

Images