Adaptive load balancing and emergency power supply system for multimedia conference equipment

Abstract

This invention relates to the field of circuit devices for power supply or distribution, and discloses an adaptive load balancing and emergency power supply system for multimedia conferencing equipment, including: a command monitoring unit, a current prediction unit, a main power supply circuit, and an adaptive controller. The line impedance observer calculates the real-time transmission resistance using the ratio of the transient value of the bus voltage to the expected increment value of the current during load step transients; the virtual impedance compensation module generates a negative impedance coefficient and reconstructs the voltage loop accordingly. This invention identifies the line impedance online and synthesizes a virtual negative resistance through command excitation, automatically offsetting the transmission voltage drop and cable temperature rise drift under the condition of no remote sampling line, ensuring the dynamic rigidity and life-cycle stability of the end voltage under existing lines.

Description

Technical Field

[0001] This invention relates to an adaptive load balancing and emergency power supply system for multimedia conferencing equipment, belonging to the technical field of power supply or distribution circuit devices. Background Technology

[0002] Current multimedia conferencing systems often employ centralized DC or AC power supply architectures to drive heterogeneous load clusters of LED display arrays, digital power amplifiers, and logic processing units. They are configured with voltage-based negative feedback regulated power supplies that sample the bus voltage in real time and compare it with a reference value. The error signal is used to adjust the duty cycle of power devices to maintain a stable output level. The closed-loop control mode maintains power quality and system reliability under steady-state conditions with smooth load changes. However, in multimedia conferencing scenarios, the instantaneous switching of video content and sudden peaks in audio signals cause the load current to have extremely high change rate and randomness. Due to the physical inertia of the power circuit and the limited bandwidth of the feedback loop, the timing of error-triggered adjustment lags behind the sudden change in load current. When the load current experiences a large step jump, the power supply system fails to complete the response adjustment. The DC bus voltage drops or oscillates due to line impedance voltage drop and output capacitor discharge.

[0003] In the field of power supply for multimedia conferencing equipment, existing technical solutions, such as the bidirectional stepped power supply circuit disclosed in Chinese patent application CN117375418A, achieve energy conversion and voltage regulation between different power supply terminals through the cooperation of main and auxiliary control units. However, such solutions still have technical bottlenecks in practical applications: their voltage stabilization mechanism is essentially still a negative feedback control triggered by deviation. Due to the physical inertia of the power circuit and the limited bandwidth of the feedback loop, the adjustment action always lags behind the sudden change of load current. When facing high dynamic pulse loads such as LED display arrays or digital power amplifiers, the DC bus voltage will experience transient drops due to the adjustment lag. Secondly, this solution does not consider the dynamic drift of the distributed impedance of long-distance transmission cables due to environmental temperature rise or connector aging, and lacks effective online impedance identification and in-situ compensation methods, making it difficult to guarantee the consistency of the input voltage of the end load. In addition, under the condition of sudden load power unloading, due to the lack of predictable energy recovery and voltage clamping mechanisms, the line inductance freewheeling is very likely to cause bus voltage overshoot, posing an electrical damage risk to low-voltage sensitive devices connected on the same network.

[0004] Therefore, how to overcome the physical lag bottleneck of traditional feedback control and achieve zero-delay voltage compensation and adaptive line impedance calibration for high dynamic heterogeneous loads without adding passive filter components and remote sampling cables has become the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: An adaptive load balancing and emergency power supply system for multimedia conferencing equipment, comprising: The instruction monitoring unit is connected to the control bus of the multimedia conferencing system and is used to collect control instruction frames sent to the load device in real time and parse out the preset key power consumption instructions. The current prediction unit, which is connected to the command monitoring unit, is used to query a preset power consumption data table based on the key power consumption command in order to determine the expected increment of the load current when the load device executes the key power consumption command. The main power supply circuit includes an adjustable voltage source for outputting a reference DC voltage to the DC bus; An adaptive controller is connected to both the current prediction unit and the main power supply loop, and integrates a line impedance observer and a virtual impedance compensation module. The line impedance observer executes impedance calculation logic, synchronously acquiring the voltage transient value at the DC bus output port during the transient period when the main power supply loop generates a step change in load current in response to a critical power consumption command, and calculating the real-time equivalent transmission resistance of the distribution network based on the ratio of the voltage transient value to the expected increment of the load current. The virtual impedance compensation module executes voltage loop regulation logic, generates a negative impedance compensation coefficient based on the real-time equivalent transmission resistance, and introduces the negative impedance compensation coefficient into the voltage reference control loop of the main power supply loop to establish a virtual negative output impedance at the output of the main power supply loop to offset the transmission voltage drop of the distribution network.

[0006] Preferably, the line impedance observer determines the real-time equivalent transmission resistance according to the following formula: ,in, For real-time equivalent transmission resistance, This represents the steady-state DC bus voltage value before the load current step occurs. This represents the extreme value of the DC bus voltage drop after a load current step change. The expected increment of the load current; the adaptive controller only updates the negative impedance compensation coefficient in the virtual impedance compensation module when the rate of change of the real-time equivalent transmission resistance calculated over multiple consecutive sampling periods is lower than the preset convergence threshold.

[0007] Preferably, it further includes: a hybrid energy storage compensation branch connected in parallel to the DC bus, comprising a supercapacitor module and a bidirectional DC-DC converter connected in series; the adaptive controller is also used to identify load unloading commands in critical power consumption commands, and send an energy recovery trigger signal to the bidirectional DC-DC converter when the load unloading command is identified; the bidirectional DC-DC converter is used to switch its control mode to synchronous rectification state and lower the reference voltage threshold on the DC bus side in response to the energy recovery trigger signal, so as to establish a low input impedance path and import the excess power on the DC bus into the supercapacitor module when the load device performs unloading action causing a sudden rise in DC bus voltage.

[0008] Preferably, the adaptive controller is also used to execute online energy storage status detection logic: when the adaptive controller determines that the load device is in a steady-state condition, it controls the bidirectional DC-DC converter to inject a small-signal disturbance current of preset frequency and amplitude into the supercapacitor module; the adaptive controller collects the voltage response signal generated by the supercapacitor module in response to the small-signal disturbance current, and uses a synchronous demodulation method to extract the component with the same frequency as the small-signal disturbance current from the voltage response signal to calculate the real-time equivalent series resistance value of the supercapacitor module; the adaptive controller dynamically adjusts the feedforward adjustment advance time of the main power supply circuit in response to subsequent key power consumption commands according to the increase in the real-time equivalent series resistance value.

[0009] Preferably, the adaptive controller further includes an instruction verification module. The instruction verification module is used to verify whether the request parameters of the key power consumption instruction exceed the preset rated operating range and whether the instruction sending frequency exceeds the preset timing threshold before the current prediction unit determines the expected incremental value of the load current. When the instruction verification module outputs an abnormal result, the adaptive controller executes a protection strategy. The protection strategy includes blocking the voltage compensation operation corresponding to the key power consumption instruction or sending a current limiting control signal to the corresponding output port to limit the power output of the main power supply circuit.

[0010] Preferably, the virtual impedance compensation module is also used to execute nonlinear gain adjustment logic: the virtual impedance compensation module has a pre-stored nonlinear compensation curve that varies with the load current amplitude; when the expected increment of the load current is in the light load range, the virtual impedance compensation module uses a linear gain mode to generate a negative impedance compensation coefficient; when the expected increment of the load current is in the heavy load range, the virtual impedance compensation module introduces a quadratic term compensation component based on the nonlinear compensation curve to compensate for the increase in nonlinear impedance caused by the temperature rise of the power distribution network cables under high current.

[0011] Preferably, the system also includes an emergency management module, which is used to monitor the remaining power of the hybrid energy storage compensation branch in real time when the external mains power is disconnected and the system enters the emergency power supply mode; the emergency management module calculates the expected energy consumption based on the currently received key power consumption instructions and estimates the remaining battery life of the system; when the remaining battery life is less than the preset meeting end time threshold, the emergency management module generates a suppression signal to intercept key power consumption instructions of high power consumption category, or forces the main power supply circuit to switch to low power output mode.

[0012] Preferably, the adjustable voltage source in the main power supply circuit includes a multi-phase parallel buck converter; the adaptive controller is also used to dynamically adjust the number of operating phases of the buck converter according to the magnitude of the expected increment of the load current; when the expected increment of the load current is greater than the preset transient threshold, the adaptive controller forces all phases of the buck converter to start and lock them in the maximum duty cycle state before the load current step occurs, so as to improve the dynamic response bandwidth of the main power supply circuit.

[0013] Preferably, the current prediction unit has parameter self-calibration logic: within the time window when the DC bus is in a steady state and no new instructions are issued, the current prediction unit obtains the actual total current value of the DC bus; the current prediction unit calculates the steady-state deviation rate between the actual total current value and the sum of the theoretical currents corresponding to each load instruction; when the steady-state deviation rate exceeds the preset dead zone range, the current prediction unit generates a correction factor based on the steady-state deviation rate and uses the correction factor to update the expected increment value of the load current stored in the power consumption data table.

[0014] Preferably, the system adopts a distributed power supply architecture; the main power supply circuit includes multiple power supply modules connected in parallel to the DC bus; the adaptive controller is connected to each power supply module through a digital communication bus, and synchronously sends the calculated negative impedance compensation coefficient to the local control loop of each power supply module to coordinate each power supply module to synchronously adjust its output voltage reference value.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. In adaptive load balancing and emergency power supply, by utilizing the timing coupling between the control bus and the power supply loop, the DC bus voltage reference or the output impedance state of the hybrid energy storage branch is pre-adjusted within the physical delay window of the load execution command before the load current step change, thereby constructing a power supply response phase lead characteristic. This eliminates the response lag caused by the reliance on error sampling in the traditional voltage negative feedback control loop, enabling the power supply system to maintain the dynamic rigidity of the DC bus voltage using a low-capacitance configuration when facing high-rate-of-change pulse loads. This solves the physical bottleneck of the power supply regulation bandwidth mismatch between centralized power supply architecture and the transient requirements of heterogeneous loads.

[0016] 2. Utilize the load response key power consumption command to generate a current step as the excitation signal for the physical parameters of the power distribution network. In real time, differentiate and compare the expected current increment with the actual voltage transient amplitude at the DC bus port to identify the equivalent resistance value of the transmission path online. In the voltage regulation loop of the main power supply circuit, synthesize a virtual impedance component with the same amplitude but opposite polarity as the equivalent resistance value. Based on in-situ excitation parameter adaptive compensation, automatically offset the voltage drop drift caused by long-distance cable temperature rise, connector contact oxidation, or line aging. Utilize the existing line configuration to ensure the consistency and stability of the voltage at the end load input port throughout its entire life cycle.

[0017] 3. Construct a bidirectional energy throughput control topology based on command polarity. When the load performs a shutdown or power relief action, force the connected bidirectional DC-DC converter with a supercapacitor to switch to a low input impedance synchronous rectification mode. Utilize the supercapacitor as an energy potential sink to actively absorb excess energy on the DC bus caused by the line inductance freewheeling, extending the voltage regulation control range from unidirectional anti-drop to bidirectional voltage clamping. The active energy absorption mechanism realizes braking energy recovery and utilization, physically suppressing voltage surges caused by load relief, and preventing bus voltage overshoot from causing electrical breakdown or permanent damage to low-voltage sensitive logic devices connected on the same network. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the adaptive load balancing and emergency power supply system for the multimedia conferencing equipment of the present invention. Figure 2 This is a schematic diagram of the module connection of the adaptive load balancing and emergency power supply system of the multimedia conferencing equipment of the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0020] This invention provides an adaptive load balancing and emergency power supply system for a multimedia conferencing device, including a command monitoring unit, a current prediction unit, a main power supply circuit, and an adaptive controller. The command monitoring unit is physically connected to the control bus of the multimedia conferencing system, including RS-485, DMX512, or industrial Ethernet, for real-time capture of digital control frames sent to the load device. The current prediction unit is connected to the command monitoring unit and queries preset data based on identified key power consumption commands, outputting the expected increment value of the load current. The main power supply circuit includes an adjustable voltage source, whose input is connected to a high-voltage DC source rectified from the power grid, and whose output is connected to a DC bus to provide a reference DC voltage. The adaptive controller, as the core calculation and control unit, establishes communication connections with both the current prediction unit and the main power supply circuit, and integrates line impedance within the controller. The system includes an observer and a virtual impedance compensation module. In addition, it includes a hybrid energy storage compensation branch connected in parallel to the DC bus, which contains a supercapacitor module and a bidirectional DC-DC converter connected in series. Addressing the uncontrollable transmission impedance problem caused by temperature rise in long-distance transmission cables or oxidation of connectors in existing technologies, this embodiment of the invention adopts an online identification and compensation strategy for line impedance based on command step excitation. The adaptive controller utilizes the transient process of current step generation in response to key power consumption commands from the load device as a large signal excitation to calculate the impedance characteristics of the transmission path online. The specific identification and compensation procedure is as follows: The command monitoring unit analyzes the data frames on the control bus in real time. When a preset key power consumption command is identified, such as an on command where the brightness changes from 0 to 255, the current prediction unit queries a preset power consumption data table based on the command to determine the expected increment value of the load current corresponding to the command. The line impedance observer then enters transient capture mode.

[0021] At the instant the main power supply circuit responds to the above command and generates current output, the voltage value at the DC bus output port is continuously acquired at a sampling rate of not less than 100kHz. The line impedance observer extracts the steady-state voltage value of the DC bus before the load current step occurs. In addition to the extreme value of the DC bus voltage drop after a load current step, the line impedance observer calculates the real-time equivalent transmission resistance of the distribution network based on the differential form of Ohm's law. The calculation formula is as follows: ,in, The real-time equivalent transmission resistance is expressed in ohms (Ω). The steady-state voltage of the DC bus is expressed in volts (V); the extreme voltage drop of the DC bus is expressed in volts (V). The expected increment of the load current is expressed in amperes (A). When the line impedance observer executes the transient capture procedure, the adaptive controller sets the observation trigger window according to the preset load hardware response delay parameter, so that the acquired voltage transient value matches the load current step caused by the critical power consumption command in timing. During the 50μs to 200μs buffer period after the command is parsed, the controller maintains the sampling standby state to avoid electromagnetic interference at the moment the drive circuit is turned on. After the buffer time ends, high-speed voltage monitoring with a sampling frequency of not less than 100kHz is started to extract the extreme value of the DC bus voltage drop after the load current step occurs, and combine it with the steady-state voltage value of the DC bus before the load current step occurs. The absolute value of the difference between the two is selected as the calculated voltage drop, making the real-time equivalent transmission resistance... The computational basis is anchored to the physical power conversion process.

[0022] The adaptive controller calculates the value only within a series of consecutive sampling periods. If the rate of change is lower than a preset convergence threshold, such as 5%, then the impedance change is considered to originate from the temperature rise or aging of the physical circuit, and the parameters are updated. The virtual impedance compensation module then updates the parameters based on the updated data. A negative impedance compensation coefficient is generated and introduced into the voltage reference control loop of the main power supply circuit. The controller then uses the voltage reference value of the main power supply circuit. Adjusted to: In the formula, This is the adjusted voltage reference value, in volts (V). This is the system's reference rated voltage, in volts (V). This is the measured real-time total load current, in amperes (A). The combined resistance at the power supply output is [value missing]. The virtual resistor generates a voltage boost that offsets the voltage drop on the physical line, ensuring stable voltage at the load end. When the virtual impedance compensation module executes nonlinear gain adjustment logic, it determines the initial proportion of the quadratic compensation component based on the pre-stored temperature coefficient of the cable material resistivity. It then uses real-time equivalent transmission resistance data output by the line impedance observer under different load current amplitudes for online regression analysis to fit a temperature rise characteristic curve representing the change in distribution network impedance with current thermal effect. When the expected increase in load current enters the heavy load range, it automatically switches the linear compensation in the voltage reference control loop to a nonlinear compensation that includes temperature rise characteristic correction. This increases the voltage compensation gain under high current conditions to offset the increase in physical impedance caused by cable heating. The specific logic of its nonlinear compensation is as follows: when the expected increase in load current... When the current exceeds 20A and enters the heavy load range, the virtual impedance compensation module introduces a correction term based on the self-product of the current value on the basis of the original linear compensation. The specific operation is as follows: the real-time total load current value is squared and multiplied by the preset temperature rise compensation gain coefficient (this coefficient is calibrated by the resistivity change rate of aluminum cable in the range of 25℃ to 75℃ and set to 0.00385). The resulting millivolt-level correction voltage is accumulated to the voltage reference control loop. If the line impedance observer detects that the real-time equivalent transmission resistance increases by more than 2% continuously within 120s, the temperature rise compensation gain coefficient is finely adjusted upward in steps of 0.0001 until the voltage deviation at the load end returns to within 1%, so that the dynamic deviation of the voltage at the end input port of the load equipment is maintained within the rated value of 1% under full power operation.

[0023] To address the DC bus voltage overshoot problem caused by sudden load shedding, this embodiment of the invention implements a signal-driven energy recovery procedure. The adaptive controller continuously monitors the command stream. When a load shedding command is detected, such as a brightness value returning to zero or a command to cut off a non-critical load, an energy recovery trigger signal is sent to the bidirectional DC-DC converter in the hybrid energy storage compensation branch within a preset buffer time before the actual load shutdown action. This preset buffer time is set to 50μs to 200μs. In response to this signal, the bidirectional DC-DC converter switches its control mode from standby or discharge state to synchronous rectification state and temporarily lowers the reference voltage threshold on the DC bus side, for example, by 0.5V. When the actual load shedding causes the bus voltage to rise due to line inductance freewheeling and exceed the lowered reference voltage threshold, the bidirectional DC-DC converter exhibits low input impedance characteristics, guiding excess energy on the DC bus into the supercapacitor module for storage, thus clamping down on voltage fluctuations. To address the degradation of dynamic response capability caused by the aging of energy storage components, this embodiment of the invention is equipped with online energy storage status detection logic. The adaptive controller monitors the control bus, and when it is determined that the system has not received a critical power consumption command within a preset time window, such as 5 seconds, i.e., it is in a steady-state condition, it controls the bidirectional DC-DC converter to inject a small-signal disturbance current with a preset frequency and amplitude into the supercapacitor module. The preset frequency is set to 1kHz, and the preset amplitude is set to 1% of the rated current. The adaptive controller synchronously collects the voltage response signal of the supercapacitor module to the small-signal disturbance current, and uses a synchronous demodulation method to extract the voltage component with the same frequency as the small-signal disturbance current, and calculates the real-time equivalent series resistance value of the supercapacitor module. When it is detected that the real-time equivalent series resistance value has increased relative to the factory reference value, the adaptive controller automatically increases the feedforward adjustment advance time of the main power supply circuit in response to subsequent critical power consumption commands to compensate for the energy release delay caused by the increase in internal resistance.

[0024] To ensure the long-term accuracy of the power consumption data table, the system executes parameter self-calibration logic. During the time window when the DC bus is in a steady state and no new instructions are issued, the current prediction unit obtains the actual total current value of the DC bus. Meanwhile, the current prediction unit calculates the theoretical total current based on the theoretical values ​​corresponding to each load command. The system calculates the steady-state deviation rate. : In the formula, δ is the steady-state deviation rate, which is dimensionless; This is the actual total current value, in amperes (A). The theoretical total current is expressed in amperes (A). When δ exceeds the preset dead zone range, for example, ±2%, the current prediction unit generates a correction factor based on δ and uses this correction factor to update the expected increment value of the load current stored in the power consumption data table, thereby enabling the model to adaptively track the power consumption drift caused by equipment aging. When establishing the power consumption data table, for load devices with continuous adjustment characteristics, multiple discrete calibration points of brightness or volume adjustment parameters in the control command are selected for sampling. The steady-state current increment value corresponding to the calibration point is measured and stored in non-volatile memory. A linear interpolation method is used to determine the continuous prediction curve over the full range of parameters. During system operation, the actual total current value of the DC bus is obtained through the current prediction unit. The steady-state deviation rate between the actual total current value and the theoretical total current corresponding to each load command is calculated. Based on the deviation rate, a correction factor is generated to dynamically update the expected increment value of the load current stored in the power consumption data table, thereby offsetting the power consumption characteristic drift caused by individual differences in equipment or component aging.

[0025] Regarding the hardware configuration of the main power supply circuit, the adjustable voltage source adopts a multi-phase parallel buck converter. The adaptive controller dynamically adjusts the number of operating phases of the converter according to the magnitude of the expected increment of the load current. When the expected increment of the load current is greater than the preset transient threshold, such as 30A, the adaptive controller forces all phases of the buck converter to start and lock them in the maximum duty cycle state before the load current step occurs, so as to improve the dynamic response bandwidth of the main power supply circuit. In addition, the adaptive controller also includes an instruction verification module to defend against logic layer attacks. Before the current prediction unit determines the expected increment of the load current, the instruction verification module verifies whether the request parameters of the key power consumption instruction exceed the preset rated operating range and whether the instruction sending frequency exceeds the preset timing threshold, such as 10 times per second. When the verification result is abnormal, the adaptive controller executes protection strategies, including blocking the voltage compensation operation corresponding to the key power consumption instruction, or sending a current limiting control signal to the corresponding output port to limit the power output of the main power supply circuit. In this embodiment, the establishment of the power consumption data table follows the following calibration procedure: at the initial system... In the calibration phase, each load device is controlled to perform key functions, such as full white screen display or maximum volume output. The steady-state current increment after each command is executed is recorded using a current sensor, and the command identifier code and the corresponding current increment value are stored in non-volatile memory to form an initial mapping table. The specific calibration steps are as follows: In the initial self-test phase of the system's first startup, the adaptive controller controls the main power supply circuit to issue step test commands at five brightness characteristic points: 0%, 25%, 50%, 75%, and 100%. After stabilizing for 500ms at each point, the steady-state current of the DC bus is collected by the current sensor. The controller writes these five discrete current values ​​and their corresponding command identifier codes into the non-volatile memory as anchor data. When the key power consumption command in actual operation is located between two anchor points, the controller obtains the difference between the current command value and the command value of the lower anchor point, multiplies it by the linear slope of the interval (i.e., the current difference between two adjacent anchor points divided by the command difference), and adds the result to the current value of the lower anchor point, thereby synthesizing a continuous expected increment value of the load current in real time.

[0026] Example 1: This example is verified in a multimedia conference center application scenario. In this scenario, the power supply network is connected to an LED display array with a rated power of 15kW, and the transmission cable length exceeds 100m. Due to long-term operation and connector oxidation, the total equivalent transmission resistance of the power distribution network is approximately 0.15Ω. When the command monitoring unit captures the start command from the control bus indicating that the brightness has changed from 0 to 255, the current prediction unit queries the power consumption data table and determines the expected increment of the load current corresponding to the command. At a current of 300A, without compensation, this current step will generate a 45V voltage drop on the transmission line, causing the load input voltage to fall below the normal operating threshold. In response to this critical power consumption command, the system initiates transient capture logic in the line impedance observer. Within a microsecond window of the main power supply circuit responding to the command and outputting current, the line impedance observer acquires the steady-state voltage value at the DC bus output port. 380V, and the extreme voltage drop value. The voltage is 335V. The line impedance observer calculates the real-time equivalent transmission resistance using the aforementioned formula. The impedance is 0.15Ω. The virtual impedance compensation module then generates a negative impedance compensation coefficient based on this impedance value and synthesizes a virtual resistor with a resistance of -0.15Ω in the voltage control loop of the main power supply circuit. As the actual load current climbs to 300A, the main power supply circuit adjusts the voltage based on the reconstructed reference value. It automatically boosts the output voltage to 45V, and this voltage boost precisely offsets the voltage drop on the physical circuit, keeping the voltage at the load end at the rated level.

[0027] When the system executes the command to disconnect non-critical loads, the load current returns to zero from 300A within 100μs. After the adaptive controller recognizes the load unloading command, it sends an energy recovery trigger signal to the hybrid energy storage compensation branch within a preset buffer time before the load actually performs the shutdown action. The bidirectional DC-DC converter responds to the signal by switching to synchronous rectification state and lowering the reference voltage threshold on the DC bus side by 0.5V. When the line inductance freewheeling causes the bus voltage to rise and exceed the lowered threshold, the bidirectional DC-DC converter exhibits low input impedance characteristics, guiding the excess power on the bus into the supercapacitor module, thus achieving voltage overshoot clamping and energy recovery.

[0028] Example 2: This example uses physical verification experiments to objectively demonstrate the effectiveness and stability of the technical solution of the present invention in simulating a real industrial environment. The experimental design not only verifies the system's ability to compensate for line impedance drift, but also evaluates the performance of the solution in suppressing load shedding overshoot and coping with equipment aging by introducing simulated noise and operating condition disturbances. The test platform consists of a programmable high-power DC power supply (simulating rectified input source, output range 0-600V, accuracy ±0.1V), an FPGA-based digital controller (running the adaptive control algorithm of the present invention), and a variable impedance network (simulating power distribution cables and contact resistance changes). The load end uses a programmable electronic load (simulating high-frequency pulse load, maximum current change rate 10A / μs) connected in parallel with a set of actual LED modules. In addition, to simulate a real industrial electromagnetic environment, Gaussian white noise with a signal-to-noise ratio of 20dB is actively superimposed in the signal sampling circuit, and power frequency interference with a frequency of 50Hz and an amplitude of 1V is introduced. All voltage and current data are collected by a high-precision power analyzer (sampling rate 2MHz, bandwidth 5MHz).

[0029] The experiment was conducted in three groups to verify the online impedance identification and virtual resistance compensation mechanism: the control group used the traditional constant voltage control mode (without compensation); the experimental group of this invention used the control mode with online identification and virtual impedance compensation enabled; and the partially missing control group used only fixed feedforward compensation (preset fixed impedance value, not updated over time). During the experiment, the variable impedance network was controlled to simulate the process of the cable impedance gradually increasing from 0.1Ω to 0.3Ω to characterize line aging or temperature rise. At the same time, the electronic load performed periodic current steps from 0A to 200A to simulate the on / off switching of the display screen. The table below shows the load voltage drop amplitude of each experimental group when facing a 200A current step under different line impedance conditions. ) and recovery time ( ).

[0030] Table 1: Comparison of Load-Side Voltage Response under Different Control Strategies

[0031] Data shows that as the line impedance increases, the voltage drop in the control group deteriorates linearly, triggering undervoltage protection when the impedance reaches 0.2Ω. While some missing control groups performed reasonably well at 0.1Ω, they failed after impedance drift. In contrast, the present invention's sample group utilizes real-time calculated impedance values ​​to dynamically adjust the virtual resistance, keeping the voltage drop consistently below 2.5V. This characteristic of the system adaptively maintaining terminal voltage rigidity regardless of external impedance changes confirms the synergistic effect of online identification + virtual impedance mechanism. To verify the energy recovery effect during load sudden unloading, a comparative experiment was conducted. The load current was instantaneously cut off from 200A to 0A. The control group, relying solely on passive capacitor filtering, experienced an instantaneous overshoot of the bus voltage to 450V (rated 380V), which continued... For a time exceeding 5ms, the prototype of this invention, upon recognizing the shutdown command, triggered the bidirectional converter to enter synchronous rectification mode 100μs in advance and lowered the reference voltage by 0.5V. Actual measurement data showed that the bus voltage overshoot was clamped within 385V, and the supercapacitor module voltage increased by 0.2V, proving that excess energy was effectively recovered. Gradient verification was conducted on the setting of the key parameter feedforward adjustment advance time, with advance times set to 0μs, 50μs, 200μs, and 500μs respectively. The results showed that: at 0μs, the voltage drop improvement was not significant; within the 50μs to 200μs range, the voltage fluctuation suppression effect reached its optimal and stable state; when increased to 500μs, due to the excessively long advance boost time, unnecessary pre-overshoot of the bus voltage occurred before the actual load operation.

[0032] Example 3: This example aims to demonstrate the rationality and engineering optimization process of key parameter settings such as feedforward adjustment advance time in the technical solution of this invention through a quantitative verification system based on gradient control experiments, and to reveal the nonlinear influence of these parameters on the overall system performance, thereby eliminating potential parameter black box problems. The experiment was conducted on a controlled simulation platform built on MATLAB / Simulink and configured with the same circuit topology parameters as in the above examples. The simulation model integrates a dynamic line impedance change module and a random pulse load generator to reproduce the complex electrical environment in a multimedia conferencing system. To determine the optimal range of feedforward adjustment advance time, a gradient control experiment was designed. While keeping other control parameters unchanged, the advance time was adjusted... As the sole variable, the values ​​were set to 0μs, 20μs, 50μs, 100μs, 200μs, 300μs, and 500μs respectively. For each set value, the system performed multiple load current step tests with an amplitude of 200A and recorded the voltage drop amplitude at the load end. and bus voltage overshoot peak The table below shows the system voltage response data under different advance time settings. All data are averages from multiple tests.

[0033] Table 2: Influence of Feedforward Regulation Advance Time on Voltage Response

[0034] Data analysis revealed a non-linear relationship between lead time and system performance, when As the voltage sag increases from 0 μs to 50 μs, the voltage drop amplitude decreases, indicating that the advance adjustment effectively compensates for the inherent delay of the control loop. Within the 50 μs to 200 μs range, the voltage sag remains at a low level of 2.1V to 2.8V, and the bus voltage overshoot does not exceed 3% of the rated value. This range is confirmed as the optimal operating window for the system. When the voltage is further increased to 300μs and above, although the dropout suppression remains effective, the bus voltage overshoot increases. This is because the premature boost operation causes the bus voltage to accumulate too much energy before the load current is established, resulting in a pre-overshoot phenomenon. This proves that the parameter selection is not arbitrary, but based on a precise trade-off between response speed and voltage safety. In addition, the timing threshold parameter in the instruction verification module was verified by constructing a malicious instruction injection model. Repeated instruction sequences with frequencies increasing from 1Hz to 100Hz were injected into the control bus. The results show that when the instruction frequency is below 10Hz, the system can respond normally and maintain stability. When the frequency exceeds 10Hz and approaches the power loop bandwidth limit, the control group without verification function showed obvious bus voltage oscillation, while the sample of this invention with verification function activated the current limiting strategy after detecting that the frequency exceeded the 10Hz threshold, forcibly clamping the instruction execution frequency within a safe range, effectively avoiding the risk of physical resonance. This result provides a clear stability boundary basis for setting the timing threshold of 10 times per second.

[0035] Example 4: Based on the verification of the power consumption data table and virtual impedance compensation module in the previous examples, this example constructs a standardized pre-deployment calibration procedure to ensure the accuracy of parameters and control stability of the system when facing different physical line lengths and load configurations. It is suitable for the initialization phase after the system's first power-on or after a physical change in the power distribution network. The static calibration logic of the line impedance is initiated, and the system disconnects all non-critical loads, retaining only one calibration load with known power characteristics, such as a purely resistive dummy load. The adaptive controller controls the main power supply circuit to output a low-amplitude probe current step, with an amplitude of 5% of the rated current. The controller synchronously collects the voltage change at the DC bus output terminal. And combined with known probe current increments The reference equivalent resistance of the power distribution network is initially calculated based on Ohm's law. This step is performed at low power to avoid potential safety risks from full-power testing and to provide safe initial parameters for subsequent dynamic compensation.

[0036] Secondly, the power consumption model dynamic correction logic is executed. After confirming the safety of the reference impedance, the system activates each actual load device and records the actual current consumption value of each device during steady-state operation. The controller compares the measured current value with the preset theoretical value in the power consumption data table and calculates the deviation coefficient. If the deviation coefficient exceeds the preset range, such as ±10%, the corresponding entry in the mapping table is automatically updated to correct the theoretical value to the measured value. This eliminates prediction errors caused by individual equipment differences or aging, ensuring the accuracy of the current prediction unit's output. To maintain a high degree of consistency with actual physical requirements, after completing the above two calibration steps, the system automatically generates and solidifies the final set of control parameters and enters normal operation mode.

[0037] Example 5: This example establishes a standardized field debugging procedure for timing threshold parameters in the instruction verification module. This ensures the system's interception effectiveness and misjudgment avoidance capability when facing different control bus protocols and load response characteristics. This procedure is applicable to the adaptive adjustment phase after the system connects to new control devices or upgrades firmware. It executes benchmark communication stress test logic, placing the instruction verification module in monitoring mode. In this mode, only the instruction frequency is recorded without blocking. A control bus analyzer is used to inject the worst-case normal instruction stream, consistent with actual application scenarios, into the system. For example, simulating high-frequency synchronous flashing instructions of multiple LED lights under musical rhythm. The adaptive controller statistically analyzes the peak arrival rate of instructions per unit time. Record the voltage fluctuation of the main power supply circuit at this rate. If the voltage fluctuation amplitude is less than the rated voltage... Then the The initial timing threshold is set to 1.2 times the value. .

[0038] Secondly, execute the abnormal boundary detection and threshold solidification logic. In monitoring mode, gradually increase the repetition frequency of the injected commands until the power loop bandwidth limit is reached or the bus voltage ripple is observed to exceed the safety standard, such as exceeding the rated value by 3%. Record the critical frequency at this point. The system will ultimately set the time-series threshold. Set as and The smaller value in the range ensures both the right to pass instructions under extreme normal operating conditions and effective interception before the physical stability boundary is reached, thus achieving a balance between security protection and business continuity.

[0039] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0040] Finally, 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. An adaptive load balancing and emergency power supply system for a multimedia conferencing device, characterized in that, include: The instruction monitoring unit is connected to the control bus of the multimedia conferencing system and is used to collect control instruction frames sent to the load device in real time and parse out the preset key power consumption instructions. The current prediction unit, which is connected to the command monitoring unit, is used to query a preset power consumption data table based on the key power consumption command in order to determine the expected increment of the load current when the load device executes the key power consumption command. The main power supply circuit includes an adjustable voltage source for outputting a reference DC voltage to the DC bus; An adaptive controller is connected to both the current prediction unit and the main power supply loop, and integrates a line impedance observer and a virtual impedance compensation module. The line impedance observer executes impedance calculation logic, synchronously acquiring the voltage transient value at the DC bus output port during the transient period when the main power supply loop generates a step change in load current in response to a critical power consumption command, and calculating the real-time equivalent transmission resistance of the distribution network based on the ratio of the voltage transient value to the expected increment of the load current. The virtual impedance compensation module executes voltage loop regulation logic, generates a negative impedance compensation coefficient based on the real-time equivalent transmission resistance, and introduces the negative impedance compensation coefficient into the voltage reference control loop of the main power supply loop to establish a virtual negative output impedance at the output of the main power supply loop to offset the transmission voltage drop of the distribution network.

2. The adaptive load balancing and emergency power supply system for a multimedia conferencing device according to claim 1, characterized in that, The line impedance observer determines the real-time equivalent transmission resistance using the following formula: ,in, For real-time equivalent transmission resistance, This represents the steady-state DC bus voltage value before the load current step occurs. This represents the extreme value of the DC bus voltage drop after a load current step change. The expected increment of the load current; the adaptive controller only updates the negative impedance compensation coefficient in the virtual impedance compensation module when the rate of change of the real-time equivalent transmission resistance calculated over multiple consecutive sampling periods is lower than the preset convergence threshold.

3. The adaptive load balancing and emergency power supply system for a multimedia conferencing device according to claim 1, characterized in that, Also includes: The hybrid energy storage compensation branch is connected in parallel to the DC bus and includes a series-connected supercapacitor module and a bidirectional DC-DC converter. The adaptive controller is also used to identify load unloading instructions in critical power consumption commands and send an energy recovery trigger signal to the bidirectional DC-DC converter when the load unloading instruction is identified. The bidirectional DC-DC converter is used to switch its control mode to synchronous rectification state and lower the reference voltage threshold on the DC bus side in response to the energy recovery trigger signal. This is to establish a low input impedance path and import the excess power on the DC bus into the supercapacitor module when the load device performs an unloading action that causes a sudden rise in the DC bus voltage.

4. The adaptive load balancing and emergency power supply system for a multimedia conferencing device according to claim 3, characterized in that, The adaptive controller is also used to execute the online detection logic of energy storage status: when the adaptive controller determines that the load device is in a steady state, it controls the bidirectional DC-DC converter to inject a small signal disturbance current of preset frequency and amplitude into the supercapacitor module. The adaptive controller acquires the voltage response signal generated by the supercapacitor module in response to small-signal disturbance current, and uses a synchronous demodulation method to extract the component with the same frequency as the small-signal disturbance current from the voltage response signal to calculate the real-time equivalent series resistance value of the supercapacitor module. The adaptive controller dynamically adjusts the feedforward adjustment advance time of the main power supply circuit in response to subsequent key power consumption commands based on the increase in the real-time equivalent series resistance value.

5. The adaptive load balancing and emergency power supply system for a multimedia conferencing device according to claim 1, characterized in that, The adaptive controller also includes an instruction verification module, which verifies whether the request parameters of the key power consumption instruction exceed the preset rated operating range and whether the instruction sending frequency exceeds the preset timing threshold before the current prediction unit determines the expected incremental value of the load current. When the instruction verification module outputs an abnormal result, the adaptive controller executes a protection strategy. The protection strategy includes blocking the voltage compensation operation corresponding to the critical power consumption instruction, or sending a current limiting control signal to the corresponding output port to limit the power output of the main power supply circuit.

6. The adaptive load balancing and emergency power supply system for a multimedia conferencing device according to claim 1, characterized in that, The virtual impedance compensation module is also used to execute nonlinear gain adjustment logic: the virtual impedance compensation module has a pre-stored nonlinear compensation curve that varies with the load current amplitude; when the expected increase in load current is in the light load range, the virtual impedance compensation module uses linear gain mode to generate a negative impedance compensation coefficient; when the expected increase in load current is in the heavy load range, the virtual impedance compensation module introduces a quadratic term compensation component based on the nonlinear compensation curve to compensate for the increase in nonlinear impedance caused by the temperature rise of the power distribution network cables under high current.

7. The adaptive load balancing and emergency power supply system for a multimedia conferencing device according to claim 3, characterized in that, The system also includes an emergency management module, which monitors the remaining power of the hybrid energy storage compensation branch in real time when the external mains power is disconnected and the system enters emergency power supply mode. The emergency management module calculates the expected energy consumption based on the currently received critical power consumption commands and estimates the remaining battery life of the system. When the remaining battery life is less than the preset meeting end time threshold, the emergency management module generates a suppression signal to intercept critical power consumption commands of high power consumption category, or forces the main power supply circuit to switch to low power output mode.

8. The adaptive load balancing and emergency power supply system for a multimedia conferencing device according to claim 1, characterized in that, The adjustable voltage source in the main power supply circuit includes a multi-phase parallel buck converter; the adaptive controller is also used to dynamically adjust the number of operating phases of the buck converter according to the magnitude of the expected increment of the load current; when the expected increment of the load current is greater than the preset transient threshold, the adaptive controller forces all phases of the buck converter to start and locks them in the maximum duty cycle state before the load current step occurs.

9. The adaptive load balancing and emergency power supply system for a multimedia conferencing device according to claim 1, characterized in that, The current prediction unit has parameter self-calibration logic: within the time window when the DC bus is in a steady state and no new instructions are issued, the current prediction unit obtains the actual total current value of the DC bus; the current prediction unit calculates the steady-state deviation rate between the actual total current value and the sum of the theoretical currents corresponding to each load instruction; when the steady-state deviation rate exceeds the preset dead zone range, the current prediction unit generates a correction factor based on the steady-state deviation rate and uses the correction factor to update the expected increment value of the load current stored in the power consumption data table.

10. The adaptive load balancing and emergency power supply system for a multimedia conferencing device according to claim 1, characterized in that, The system adopts a distributed power supply architecture; the main power supply circuit includes multiple power modules connected in parallel to the DC bus; The adaptive controller connects to each power module via a digital communication bus and synchronously sends the calculated negative impedance compensation coefficient to the local control loop of each power module to coordinate the synchronous adjustment of the output voltage reference value of each power module.

Images