Computer fan grading regulation control system based on multi-sensor feedback

Abstract

The invention discloses a computer fan grading regulation control system based on multi-sensor feedback, which relates to the technical field of fan grading regulation, takes a fan grading regulation control platform as a core, and cooperatively controls an execution decision unit, a regulation execution influence evaluation unit and a fault-tolerant analysis substitution unit. An internal hardware performance evaluation unit and an operation control unit are combined to form a closed-loop control link of perception-decision-execution-feedback; the aging coefficient dynamic evaluation mechanism is used for acquiring the operation deviation and data precision difference of the control main body and the monitoring main body in a continuous / single adjustment scene in real time, constructing an aging coefficient curve after de-dimensionalization processing, and accurately quantifying the hardware performance attenuation trend; the environment self-adaptive adjusting strategy is based on the matching relation between environment parameters such as air pressure and the graded adjusting wind speed range; and fault-tolerant analysis and anomaly suppression capability are realized, and the reliability of an adjustment instruction is ensured through monitoring data anomaly detection and accidental data discrimination.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fan step adjustment, in particular to a computer fan step adjustment control system based on multi-sensor feedback. BACKGROUND

[0002] In the field of computer heat dissipation control, the traditional fan adjustment system usually adopts fixed threshold triggering or simple linear adjustment mode, which is difficult to cope with the double challenges of complex environmental changes and hardware aging.

[0003] In the prior art, there is a lack of dynamic evaluation mechanism between the fan control platform and the execution unit, resulting in problems such as adjustment delay and wind speed inaccuracy when the hardware performance declines; at the same time, environmental factors such as air pressure changes will significantly affect the matching degree of the preset wind speed and the actual heat dissipation effect, and the system lacks the ability to quantitatively evaluate the adjustment effect under non-ideal working conditions.

[0004] In addition, when the sensor data is abnormal, the system often directly interrupts the adjustment or uses a fixed replacement value, which not only affects the continuity of heat dissipation but also may exacerbate hardware wear and tear. That is, in the current system architecture, the control decision, impact evaluation and fault tolerance processing modules are isolated from each other, and cannot realize the cooperative optimization of hardware state monitoring, environmental parameter compensation and abnormal data processing. Therefore, the present application proposes a solution. SUMMARY

[0005] The purpose of the present application is to solve the above-mentioned problems, and a computer fan step adjustment control system based on multi-sensor feedback is proposed.

[0006] The purpose of the present application can be achieved by the following technical solution: a computer fan step adjustment control system based on multi-sensor feedback, comprising a fan step adjustment control platform, and a control execution decision unit, a regulation execution impact evaluation unit and a fault tolerance analysis replacement unit in communication connection with the fan step adjustment control platform; The control execution decision unit is further in communication connection with an internal hardware performance evaluation unit and a running control unit; The fan step adjustment control platform is used to generate control execution decision signals, regulation execution impact evaluation signals and fault tolerance analysis replacement signals; The control execution decision unit is used to receive the control execution decision signals, and obtain the aging coefficients of the control subject and the monitoring subject through the internal hardware performance evaluation unit, and perform execution decision analysis based on the aging coefficient curve; The regulation execution impact evaluation unit is used to receive the regulation execution impact evaluation signals, and perform impact evaluation on the execution effect of the regulation control instruction of the current control subject; The fault tolerance analysis replacement unit is used to receive the fault tolerance analysis replacement signals, and perform fault tolerance analysis and replacement control on the abnormal data in the fan step adjustment process; The internal hardware performance evaluation unit is used to perform aging performance tests on the control unit and the monitoring unit. The operation control unit is used to perform control operations such as hardware maintenance and operation intensity adjustment.

[0007] As a preferred embodiment of the present invention, the method for calculating the aging coefficient of the controlled body includes: During the cumulative usage period, in the continuous adjustment scenario, the deviation value between the adjustment buffer time of adjacent levels of the control body and the set adjustment command interval time is obtained; in the single adjustment scenario, the cumulative operating deviation between the start and stop time of the control body and the execution time of the set command is obtained; the deviation values ​​are numerically extracted and the unit influence is excluded before summing to obtain the aging coefficient of the control body.

[0008] As a preferred embodiment of the present invention, the method for calculating the aging coefficient of the monitored subject includes: During the cumulative usage period, in the continuous adjustment scenario, the data generation of the control subject is obtained during the time interval between adjacent set monitoring cycles of the monitoring subject. In the single adjustment scenario, the accuracy deviation between the accuracy of the monitoring subject on the monitoring data in the current monitoring cycle and the accuracy of the corresponding data value of the control subject is obtained. The data generation and accuracy deviation are dimensionless and the influence of units is eliminated before summing to obtain the aging coefficient of the monitoring subject.

[0009] In a preferred embodiment of the present invention, the execution decision analysis method of the control execution decision unit includes: Record the fluctuation stages of the aging coefficient of the main body and the aging coefficient of the main body within the cumulative usage period and construct the aging coefficient curve; If both are stable trends, then hardware feedback is provided based on the adjustment and control commands; If the aging coefficient of the control unit fluctuates while the monitoring unit remains stable, then the control unit is maintained by analyzing the logs and adjusting the operating intensity through the running control unit before the execution command is fed back. If the aging coefficient of the control unit is stable while that of the monitoring unit is fluctuating, the control unit is used to perform quality inspection and maintenance on the monitoring unit and then re-monitor before feedback of the execution command. If both fluctuate, stop the computer's operation, immediately inspect and simultaneously maintain the control and monitoring entities, modify the historical work logs, and mark the abnormal hardware parameters.

[0010] As a preferred embodiment of the present invention, the impact assessment method of the adjustment execution impact assessment unit includes: Obtain the air pressure at the computer's location at the moment the current regulation and control command is generated and mark it as an influencing parameter; Compute the non-intersection wind speed value and its span of the control subject's hierarchical regulation wind speed range and the corresponding level set wind speed range under the current air pressure; When the air pressure floats, combine the control period below the set level wind speed, the control period above the set level wind speed, and the computer internal temperature rise span in the hierarchical regulation control stage to divide the control time length type: When the temperature rise span exceeds the set threshold, the period below the set level wind speed is an abnormal heat dissipation period, and the period above the set level wind speed is an energy-consuming heat dissipation period; When the temperature rise span does not exceed the set threshold, the period below the set level wind speed is an energy-saving heat dissipation period, and the period above the set level wind speed is a heat dissipation energy-consuming period; For the case where the non-intersection wind speed span exceeds the set span threshold, the following strategy is used for control: When the abnormal heat dissipation period ratio rises, the level span of hierarchical regulation is increased; When the energy-consuming heat dissipation period ratio rises, the fan level is lowered when the temperature rise is controllable; When the energy-saving heat dissipation period ratio rises, the hierarchical regulation is suspended; When the energy-consuming heat dissipation period ratio rises, the fan level is continuously lowered according to the numerical ratio of the heat dissipation speed to the temperature rise speed.

[0011] As a preferred embodiment of the present application, the fault-tolerant control method of the fault-tolerant analysis replacement unit comprises: When the sensor monitoring data is abnormal, the generated adjustment control instruction is fault-tolerant analysis replaced; When the sensor monitoring data is occasional data, the floating trend of the same type data of the adjacent historical time is collected, and the influence data is detected whether it is floating combined with the interval period between the adjacent historical time and the current time; if the influence data does not float and the floating trend of the same type data of the next time is synchronized with the preset actual value, it is determined that it is occasional data, the current period adjustment instruction is not executed, and the data is marked as occasional data in the work log.

[0012] Compared with the prior art, the present application has the following advantages: 1. Overall, it effectively solves the problem of control precision decline caused by hardware performance degradation, and improves the stability of heat dissipation efficiency in complex environment; by introducing fault-tolerant analysis mechanism to avoid misoperation caused by sensor occasional abnormality, prolong the service life of hardware equipment, and dynamic evaluation mechanism of environmental parameters and heat dissipation effect to realize accurate balance of energy consumption and heat dissipation demand, compared with traditional control mode, unnecessary energy consumption is reduced.

[0013] 2. From the local point of view, the aging state of the fan control main body can be accurately identified, providing reliable basis for subsequent maintenance decision, such as when the aging coefficient exceeds the threshold value, the system can trigger lubrication maintenance or component replacement operation in advance to avoid heat dissipation failure or abnormal energy consumption caused by hardware performance decline, at the same time, by distinguishing the deviation types of different adjustment scenes, the control parameters of the actuator can be optimized to prolong the service life of the hardware; It also effectively solves the problem of misjudgment caused by single scene in sensor aging evaluation, improves the accuracy and reliability of aging coefficient calculation through multi-scene data fusion and standardization processing, provides accurate basis for subsequent maintenance decision, and avoids the control instruction execution deviation caused by the performance degradation of sensor not being identified in time.

[0014] 3. From the local point of view, the adjustment instruction failure or system downtime problem caused by hardware aging misjudgment in the prior art is solved; by dynamically analyzing the floating state of the double aging coefficient curve, the performance degradation of the control main body and the monitoring main body can be accurately distinguished to avoid the chain reaction caused by single fault point, such as timely calibration when the monitoring main body drifts to prevent the issuance of incorrect adjustment instructions, and reducing the running intensity in advance when the control main body appears mechanical wear to prolong the service life of the fan, at the same time, the forced maintenance mechanism under the double floating state effectively prevents the system crash risk caused by composite failure. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to facilitate those skilled in the art to understand, the present application will be further described below in conjunction with the drawings.

[0016] Figure 1 The system principle block diagram of the present application; Figure 2 The method flow chart of the present application. DETAILED DESCRIPTION

[0017] In order to make the person skilled in the art better understand the present application scheme, the technical scheme in the embodiment of the present application will be described clearly and completely below in conjunction with the drawings in the embodiment of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor belong to the scope of protection of the present application.

[0018] Reference to an "embodiment" herein means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily mutually exclusive of one another. As will be apparent to those of ordinary skill in the art, embodiments described herein can be combined with other embodiments.

[0019] In the prior art, the computer cooling system usually adopts fixed rotating speed or simple temperature threshold adjustment mode, which is difficult to adapt to the performance degradation caused by complex environmental changes and hardware aging. The traditional control method lacks dynamic response mechanism for sensor data anomalies and hardware state changes, resulting in unstable cooling efficiency or excessive energy consumption. For example, in the scene of air pressure fluctuation or sensor occasional failure, the conventional system cannot effectively balance the cooling demand and energy consumption control, which easily causes hardware overheating or frequent start-stop problems.

[0020] In order to solve the above problems, it is necessary to build a cooling control system which can dynamically evaluate hardware performance, adjust control strategy in real time and has fault tolerance capability. Firstly, the influence of hardware aging on execution accuracy is considered, and the decision basis is established by quantifying the performance degradation of control subject and monitoring subject. Secondly, the influence of environmental parameters on cooling effect evaluation mechanism is introduced to solve the problem of wind speed control deviation caused by air pressure change. Finally, for the scene of sensor data anomaly, history trend comparison and alternative control logic need to be designed to avoid system instability caused by misoperation.

[0021] Please refer to Figure 1 The application proposes a computer fan hierarchical regulation control system based on multi-sensor feedback, which includes a fan hierarchical regulation control platform. The fan hierarchical regulation control platform is communicatively connected with a control execution decision unit, a regulation execution influence evaluation unit, and a fault tolerance analysis and replacement unit. The control execution decision unit is communicatively connected with an internal hardware performance evaluation unit and a running control unit. The fan hierarchical regulation control platform is a central module that generates and distributes control signals. It can be implemented by an embedded controller, which is used to coordinate the working sequence and data interaction of each functional unit. The control execution decision unit is a decision module that analyzes the hardware state. It can be implemented by a microprocessor with data processing capability, which determines whether the hardware state affects the control accuracy by obtaining the aging coefficient.

[0022] The adjustment execution influence evaluation unit refers to a functional module for evaluating the control effect, which can be implemented by a data fusion processing module of a pressure sensor and a temperature sensor, and is used for analyzing the influence degree of environmental parameters on the heat dissipation effect. The fault tolerance analysis replacement unit refers to a fault tolerance module for processing data exceptions, which can be implemented by a logic processor with a historical data cache function, and identifies occasional exceptions by comparing adjacent time data trends.

[0023] Specifically, when the system is working, the control platform generates a control signal to trigger the operation of each unit. After the control execution decision unit obtains the aging coefficients of the fan and the sensor, it selects the normal execution, maintenance operation or shutdown maintenance strategy according to the aging trend.

[0024] The adjustment execution influence evaluation unit collects air pressure data in real time, calculates the deviation range of the actual wind speed from the set value, divides the heat dissipation effect type in combination with the temperature change trend, and adjusts the control strategy. When the sensor has abnormal data, the fault tolerance analysis replacement unit determines whether to execute the replacement control scheme by comparing the historical data trend to determine whether it is an occasional failure. Each unit exchanges information through a data bus to form a closed-loop control logic.

[0025] Compared with the prior art, the present scheme realizes three controls of hardware state monitoring, environmental parameter compensation and exception handling through a multi-unit cooperative working mechanism. The conventional system only triggers the fan start-stop depending on the temperature threshold, and cannot identify the control delay caused by hardware aging or the decrease in sensor accuracy. The present scheme quantitatively analyzes the aging coefficient and the environmental parameter, dynamically adjusts the control strategy, effectively corrects the wind speed control deviation in the air pressure fluctuation scene, and keeps the system stable when the sensor has occasional failures.

[0026] Through the above technical scheme, the problem of control accuracy decrease caused by hardware performance degradation is effectively solved, and the stability of the heat dissipation efficiency in complex environments is improved. The fault tolerance analysis mechanism is introduced to avoid misoperation caused by occasional sensor exceptions, prolong the service life of hardware devices, and the dynamic evaluation mechanism of environmental parameters and heat dissipation effect realizes the precise balance of energy consumption and heat dissipation demand, which reduces unnecessary energy consumption compared with the traditional control method.

[0027] The fan hierarchical regulation platform generates a control execution decision signal and sends it to the control execution decision unit; After receiving the control execution decision signal, the control execution decision unit controls the computer internal fan hierarchical regulation; According to the regulation control system, the computer internal fan is set as the control subject, and the sensor set in the computer is marked as the monitoring subject; the cumulative use time length of the current computer is determined according to the running time of the real-time regulation control system; In the cumulative use duration phase, the control subject and the monitoring subject are subjected to aging performance test by an internal hardware performance evaluation unit; In the continuous adjustment scenario, the deviation value between the adjacent level adjustment buffer duration of the control subject and the set adjustment instruction interval duration is obtained, and in the single adjustment scenario, the cumulative running deviation amount is obtained, that is, the time deviation amount between the adjustment start time and the set instruction execution time, and the time deviation amount between the adjustment stop time and the set instruction execution time, with the running time as the parameter; The interval duration deviation and the cumulative running deviation are subjected to numerical extraction, and if the units are not unified, the unit influence is excluded, the numerical summation calculation is performed, and the obtained numerical value is marked as the control subject aging coefficient; In the continuous adjustment scenario, the data generation amount of the control subject in the adjacent set monitoring period interval by the monitoring subject is obtained, wherein the data generation amount is the fan running time, the speed change amount and the like; In the single adjustment scenario, the accuracy deviation between the monitoring accuracy of the monitoring data in the current monitoring period by the monitoring subject and the accuracy of the corresponding arbitrary data value accuracy of the control subject is obtained, such as the accuracy unit of 1 second or 1 millimeter and the like; The collected data generation amount and accuracy deviation are subjected to non-dimensionalization processing, and the data values are extracted after the processing, the influence of the non-unified units is excluded, the summation calculation is performed on the extracted data values, and the numerical value is marked as the monitoring subject aging coefficient; The control execution decision unit performs execution decision analysis according to the aging coefficient: The numerical floating phase of the control subject aging coefficient and the monitoring subject aging coefficient in the cumulative use duration phase is recorded, and the aging coefficient curve is constructed; According to the aging coefficient curve analysis: If the control subject aging coefficient and the monitoring subject aging coefficient are both stable trends, then when receiving the adjustment control, the hardware feedback is performed according to the adjustment control instruction, such as fan start, wind speed level continuous increase, wind speed level adaptive adjustment and the like; If the control subject aging coefficient is a floating trend, and the monitoring subject aging coefficient is a stable trend, then the use log of the control subject in the cumulative use duration phase is analyzed, the hardware of the control subject itself is maintained by the running control unit, and the running intensity of the control subject is adjusted and controlled, such as control running duration or running frequency; after the running intensity adjustment control is completed, the adjustment control instruction is fed back and executed; If the control subject aging coefficient is a stable trend, and the monitoring subject aging coefficient is a floating trend, then the monitoring subject is subjected to quality inspection and maintenance by the running control unit, and the control subject is re-monitored, after the quality inspection and maintenance and the re-monitoring are completed, the adjustment control instruction is fed back and executed; If both the control subject aging coefficient and the monitoring subject aging coefficient are floating trends, no adjustment control instruction is received, the running task of the current computer is stopped, equipment maintenance is immediately carried out, and synchronous maintenance is carried out on the control subject and the monitoring subject, and the historical working log is modified in combination with the current computer hardware performance parameters, and if there is an abnormal hardware performance parameter, the running work type or work intensity at the historical time is marked; Wherein, the cumulative use time length stage refers to the continuous or intermittent working time period of the computer fan in a specific running period, which can be realized by recording the start and stop time stamps of the fan through the system timing module and calculating the difference, for evaluating the degradation trend of hardware performance over time, the deviation value refers to the difference between the actual response time of the fan when switching between different adjustment levels and the preset standard time, which can be obtained by comparing the time difference between the sending time of the adjacent level adjustment instruction and the time when the fan speed reaches the target value, reflecting the wear degree of mechanical parts. The cumulative running deviation amount includes start deviation and stop deviation, which can be realized by calculating the sum of the difference between the actual start time and the instruction required time, and the difference between the actual stop time and the instruction required time of the fan under a single adjustment scene, for quantifying the response delay of the actuator.

[0028] Specifically, in the continuous adjustment scene, for example, the fan needs to be gradually promoted from low level to high level, the system will record the actual buffer time length of each level switching, and compare it with the preset interval time, and extract the difference value as the deviation value. In the single adjustment scene, for example, the fan only needs to perform a start-stop operation, the system will record the deviation amount of the actual start-stop time and the instruction required time, and the start deviation and the stop deviation are accumulated. After all the deviation values are normalized, the control subject aging coefficient is generated by weighting or direct summation, which can comprehensively reflect the performance degradation of the fan actuator, such as mechanical wear and circuit aging.

[0029] Compared with the prior art, the traditional method usually only judges the hardware aging based on the cumulative running time or simple threshold, while the present scheme can more accurately quantify the actual degradation degree of hardware performance by distinguishing between continuous adjustment and single adjustment scenes, combining multi-dimensional data of dynamic response deviation and static execution deviation. For example, the prior art may ignore the instantaneous response delay of the fan when frequently switching levels, while the present scheme can capture the subtle wear changes of the gear set or motor bearing by extracting the buffer time deviation.

[0030] By the technical solution, the aging state of the fan control main body can be accurately identified, and reliable basis is provided for subsequent maintenance decision, such as when the aging coefficient exceeds the threshold value, the system can trigger lubrication maintenance or component replacement operation in advance, so as to avoid heat dissipation failure or energy consumption anomaly caused by hardware performance decline, and meanwhile, by distinguishing the deviation types of different adjustment scenes, the control parameters of the actuator can be optimized in a targeted manner, and the hardware service life is prolonged. The data generation amount refers to the number of control main body operation parameters recorded by the monitoring main body in the adjacent set monitoring period interval in the continuous adjustment scene, which can be realized by adopting the cumulative values of the fan running time and the speed change amount, and is used to reflect the data acquisition capability of the monitoring main body in the dynamic adjustment environment.

[0031] The precision deviation refers to the difference between the data recorded by the monitoring main body and the actual control main body operation parameters in the single adjustment scene, which can be realized by comparing the time precision error, length precision error and other numerical differences, and is used to quantify the measurement accuracy of the monitoring main body in the steady state environment.

[0032] Specifically, in the continuous adjustment scene, the monitoring main body needs to continuously collect the fan running time and the speed change amount in the fixed interval monitoring period, and count the total data amount as the evaluation basis in the continuous adjustment scene; in the single adjustment scene, the second-level error of the fan start and stop time stamp recorded by the monitoring main body and the real execution time, or the millimeter-level error of the speed measurement value and the set value, is compared, and the precision deviation value is calculated.

[0033] The data in the two scenes are superimposed and summed after normalization, forming an aging coefficient that comprehensively reflects the performance degradation of the sensor, such as the data generation amount in the continuous adjustment scene can be normalized to a value in the interval of 0 to 1, and the precision deviation is converted to a value in the same interval through error percentage, and the final aging coefficient is obtained after weighted summation of the two.

[0034] Compared with the prior art, the sensor aging is usually evaluated only by error statistics in a single scene or data amount in a fixed period, without considering the performance difference between dynamic adjustment and steady state operation scenes, while the present scheme quantifies the data acquisition capability and measurement precision deviation in continuous adjustment and single adjustment scenes respectively, realizes comprehensive evaluation of multi-dimensional data through de-dimensioning processing, and can more comprehensively reflect the actual aging state of the monitoring main body; The above technical solution effectively solves the problem of misjudgment caused by single scene in sensor aging evaluation, improves the accuracy and reliability of aging coefficient calculation through multi-scene data fusion and standardization processing, provides accurate basis for subsequent maintenance decision, and avoids control instruction execution deviation caused by sensor performance degradation not being recognized in time; After the control execution decision unit completes the analysis, a decision execution signal is sent to the fan staged regulation control platform, and a regulation execution influence evaluation signal is generated and sent to the regulation execution influence evaluation unit; After the regulation execution influence evaluation unit receives the regulation execution influence evaluation signal, the influence of the current control subject's regulation control instruction execution is evaluated; When the current regulation control instruction is generated, the atmospheric pressure at the computer's location at the generation time is obtained, and the atmospheric pressure is marked as an influence parameter; With the influence parameter at the current generation time, non-intersection wind speed values of the control subject's staged regulation wind speed range and the corresponding level set wind speed range at the generation time are obtained, and non-intersection wind speed spans are obtained according to each non-intersection wind speed value; it needs to be explained that the non-intersection wind speed span is represented as the current wind speed being higher than the existing level or lower than the existing level; When the influence parameter at the current generation time fluctuates, the control time period below the set level wind speed and the control time period above the set level wind speed in the current staged regulation control stage are obtained; combined with the computer internal temperature rise span in the staged regulation control stage, the control time length type is divided: If the temperature rise span exceeds the set threshold, the control time period below the set level wind speed is marked as an abnormal heat dissipation period; the control time period above the set level wind speed is marked as an energy-consuming heat dissipation period; If the temperature rise span does not exceed the set threshold, the control time period below the set level wind speed is marked as an energy-saving heat dissipation period; the control time period above the set level wind speed is marked as a heat dissipation energy-consuming period; When the non-intersection wind speed span exceeds the set span threshold, an execution high influence signal is generated and sent to the fan staged regulation control platform, and influence control is performed: If the abnormal heat dissipation period ratio rises, the staged regulation in the current stage is increased in level span to overcome the pressure influence and ensure the heat dissipation efficiency; If the energy-consuming heat dissipation period ratio rises, the temperature rise in the current stage is monitored, and the corresponding fan level is lowered when the temperature rise is within the control range; If the energy-saving heat dissipation period ratio rises, the staged regulation in the current stage is suspended; If the heat dissipation energy-consuming period ratio rises, the current computer internal heat dissipation speed and temperature rise speed are compared, and the level is adjusted according to the floating trend of the numerical ratio, i.e. when the numerical ratio increases, the corresponding fan level is continuously lowered; The aging coefficient curve refers to a trend line formed by collecting performance degradation data of the control subject and the monitoring subject in the continuous use process. Specifically, the aging coefficient can be curve-fitted according to the time dimension by using a time series analysis method to quantify the hardware performance degradation degree. The stable trend refers to that the fluctuation amplitude of the aging coefficient in the preset time window does not exceed the set threshold. Specifically, the sliding window algorithm can be used to calculate the variance or standard deviation to determine whether the hardware is in a normal working state.

[0035] The fluctuation refers to that the fluctuation amplitude of the aging coefficient in the preset time window exceeds the set threshold. Specifically, a threshold comparator can be used to monitor the data change amplitude in real time to trigger a hardware abnormality warning. The use of log analysis refers to extracting maintenance events, running intensity parameters and abnormal records in the historical operation records. Specifically, a log analysis tool can be used to extract key fields and perform correlation analysis to locate the potential causes of hardware performance degradation.

[0036] Specifically, the execution decision analysis process first continuously collects running data of the control subject and the monitoring subject through the internal hardware performance evaluation unit, such as the adjustment response delay of the fan drive module and the data collection interval deviation of the sensor. The cumulative use time length stage is divided into multiple time windows, and the mean and variance of the aging coefficient in each window are calculated to generate the aging coefficient curve.

[0037] When both aging coefficient curves show a stable trend, the system maintains the execution logic of the current adjustment instruction. When the control subject aging coefficient fluctuates, the system automatically retrieves the fan start-stop record, speed adjustment frequency and running load data in this time period, and executes hardware cleaning or reduces the maximum speed threshold through the running control unit. When the monitoring subject aging coefficient fluctuates, the system performs self-checking and calibration on the sensor and re-collects data to verify the accuracy. If both fluctuate, the system immediately terminates the task process, triggers the forced maintenance mode and updates the hardware state marker in the log.

[0038] Compared with the prior art, the traditional method usually adopts fixed period maintenance or single index warning mechanism, which cannot distinguish the independent aging states of the control subject and the monitoring subject, and is prone to cause excessive maintenance or missed detection. However, the present case can accurately identify the abnormal source by constructing a double-variable aging coefficient curve and establishing a hierarchical response strategy, such as calibrating the sensor when the monitoring subject is abnormal instead of replacing the fan, thereby reducing unnecessary hardware maintenance operations.

[0039] Through the technical scheme, the problem of invalidation of adjustment instructions or system downtime caused by hardware aging misjudgment in the prior art is solved, the performance degradation of the control subject and the monitoring subject can be accurately distinguished by dynamically analyzing the floating state of the double aging coefficient curve, the chain reaction caused by a single fault point is avoided, the timely calibration when the data of the monitoring subject drifts can prevent the issuance of incorrect adjustment instructions, the running intensity is reduced in advance when the mechanical wear of the control subject occurs, the service life of the fan is prolonged, and the system collapse risk caused by a composite fault is effectively prevented by the forced maintenance mechanism under the double floating state. After the adjustment execution influence evaluation unit completes the analysis and targeted control, the fan hierarchical adjustment control platform performs adjustment control instruction execution. A fault-tolerant analysis replacement signal is generated at the same time and sent to the fault-tolerant analysis replacement unit. After the fault-tolerant analysis replacement unit receives the fault-tolerant analysis replacement signal, fault-tolerant analysis control is performed on the fan hierarchical adjustment. When the sensor monitoring data is abnormal, the fault-tolerant analysis replacement is performed on the generated adjustment control instruction. When the sensor monitoring data is occasional data, the floating trend of the same type of data of the adjacent historical time is collected, and the influence data of the same type of data is floating detected according to the interval period between the adjacent historical time and the current time, if the influence data does not float, the preset value of the corresponding same type of data is obtained according to the floating trend, and is set as the preset actual value, if the same type of data of the next time compared with the preset actual value, the corresponding floating trend and the data floating trend of the adjacent historical time are synchronized, it is inferred that the data is occasional data, and the adjustment control instruction of the current period is not executed, and the detection value of the same type of data is marked as occasional data in the work log. The occasional data refers to the non-persistent abnormal data generated by the sensor in a short time, which can be specifically identified by comparing the fluctuation range of the adjacent historical time data, for example, the time series analysis algorithm is used to extract the data change rule. The floating trend of the same type of data refers to the continuous change direction of the monitoring value of the same sensor in the historical time window, which can be specifically calculated by using the moving average method or the difference method to calculate the increasing or decreasing trend of the adjacent data points. The preset actual value refers to the predicted value derived based on the historical data, which can be specifically generated by using the linear regression model or the exponential smoothing algorithm. The interval period detection influence data refers to judging whether there is an external environmental parameter mutation between the adjacent historical time and the current time, which can be specifically achieved by auxiliary monitoring of the air pressure and temperature sensor.

[0040] Please refer to Figure 2 As shown in the figure, the application further provides a computer fan hierarchical adjustment control method based on multi-sensor feedback, comprising the following steps: S1: The fan step adjustment control platform generates a control execution decision signal and sends it to the control execution decision unit; S2: The control execution decision unit obtains the aging coefficients of the control subject and the monitoring subject through the internal hardware performance evaluation unit, performs execution decision analysis based on the aging coefficient curve, and generates a decision execution signal; S3: The adjustment execution influence evaluation unit receives the influence evaluation signal, combines the air pressure, non-intersection wind speed span and temperature rise span, evaluates the execution effect of the adjustment control instruction, and adjusts accordingly; S4: The fault tolerance analysis replacement unit receives the fault tolerance analysis replacement signal, and performs fault tolerance processing on the sensor abnormality or occasional data.

[0041] The threshold or the preset value, the preset range and the like are set for result comparison and analysis, so as to determine whether it is good or bad, and the size of the threshold is set according to the large model analysis of sample data and artificial experience, and is recorded and stored, and can be appropriately adjusted through seasonal or reasonable influence conditions. The weight proportion coefficient and the influence factor are set according to the influence of each parameter on the result, and the specific numerical value is finally reflected on the influence of the result, and the weight proportion coefficient and the influence factor are set according to the large model analysis of sample data and artificial experience, and are recorded and stored, and can be appropriately adjusted through seasonal or reasonable influence conditions.

[0042] The preferred embodiments of the application disclosed above are only used to help explain the application. The preferred embodiments do not describe all the details, and the application is not limited to the specific embodiments. Obviously, according to the content of the specification, many modifications and changes can be made. The embodiments are selected and described in the specification in order to better explain the principles and practical applications of the application, so that those skilled in the art can well understand and utilize the application. The application is limited by the claims and their entire scope and equivalents.

Claims

1. A multi-sensor feedback based computer fan step regulation control system, characterized by, The fan step regulation control platform is in communication connection with a control execution decision unit, a regulation execution influence evaluation unit and a fault tolerance analysis substitution unit; The control execution decision unit is further in communication connection with an internal hardware performance evaluation unit and a running control unit; The fan step regulation control platform is used to generate a control execution decision signal, a regulation execution influence evaluation signal and a fault tolerance analysis substitution signal; The control execution decision unit is used to receive the control execution decision signal, obtain the aging coefficients of the control subject and the monitoring subject through the internal hardware performance evaluation unit, and perform execution decision analysis based on the aging coefficient curve; The regulation execution influence evaluation unit is used to receive the regulation execution influence evaluation signal, and perform influence evaluation on the execution effect of the regulation control instruction of the current control subject; The fault tolerance analysis substitution unit is used to receive the fault tolerance analysis substitution signal, and perform fault tolerance analysis and substitution control on abnormal data in the fan step regulation process; The internal hardware performance evaluation unit is used to perform aging performance testing on the control subject and the monitoring subject; The running control unit is used to perform control operation of hardware maintenance and running strength regulation.

2. The computer fan step regulation control system based on multi-sensor feedback according to claim 1, wherein, The method for calculating the control subject aging coefficient comprises: In the cumulative use duration stage, the deviation value of the control subject adjacent grade regulation buffer duration and the set regulation instruction interval duration is obtained in the continuous regulation scene, and the cumulative running deviation amount of the control subject regulation start and stop time and the set instruction execution time is obtained in the single regulation scene; the deviation value is summed after numerical extraction and unit influence exclusion to obtain the control subject aging coefficient.

3. The computer fan step regulation control system based on multi-sensor feedback according to claim 1, wherein, The method for calculating the monitoring subject aging coefficient comprises: In the cumulative use duration stage, the data generation amount of the control subject in the adjacent set monitoring period gap period of the monitoring subject is obtained in the continuous regulation scene, and the accuracy deviation of the monitoring data accuracy of the monitoring subject in the current monitoring period and the corresponding data value accuracy of the control subject is obtained in the single regulation scene; the data generation amount and the accuracy deviation are summed after de-dimensioning processing and unit influence exclusion to obtain the monitoring subject aging coefficient.

4. The computer fan step regulation control system based on multi-sensor feedback according to claim 1, wherein, The execution decision analysis method of the control execution decision unit comprises: The numerical floating stage of the control subject aging coefficient and the monitoring subject aging coefficient in the cumulative use duration stage is recorded and the aging coefficient curve is constructed; If both are stable trends, then perform hardware feedback according to the regulation control instruction; If the control subject aging coefficient is floating and the monitoring subject is stable, then analyze the use log and maintain the control subject hardware and regulation running strength through the running control unit and then feedback the execution instruction; If the control subject aging coefficient is stable and the monitoring subject is floating, then perform quality inspection and maintenance on the monitoring subject through the running control unit and then feedback the execution instruction after re-monitoring; If both are floating, then stop the computer running task, immediately overhaul and synchronously maintain the control subject and the monitoring subject, modify the historical work log and mark the abnormal hardware parameter.

5. The computer fan step regulation control system based on multi-sensor feedback according to claim 1, wherein, The influence evaluation method of the regulation execution influence evaluation unit comprises: The atmospheric pressure of the computer location at the current regulation control instruction generation time is obtained and marked as an influence parameter; Compute the non-intersection wind speed value and its span of the control subject's step adjustment wind speed range and the corresponding level set wind speed range under the current air pressure; When the air pressure fluctuates, combine the control period below the set level wind speed, the control period above the set level wind speed, and the computer internal temperature rise span in the step adjustment control stage to divide the control time length type.

6. A multi-sensor feedback based computer fan step regulation control system as claimed in claim 5, wherein, When the temperature rise span exceeds the set threshold, the period below the set level wind speed is an abnormal heat dissipation period, and the period above the set level wind speed is an energy-consuming heat dissipation period; When the temperature rise span does not exceed the set threshold, the period below the set level wind speed is an energy-saving heat dissipation period, and the period above the set level wind speed is a heat dissipation energy-consuming period.

7. A multi-sensor feedback based computer fan step regulation control system as claimed in claim 6, wherein, When the non-intersection wind speed span exceeds the set span threshold, perform strategy control: When the abnormal heat dissipation period proportion rises, increase the level span of step adjustment; When the energy-consuming heat dissipation period proportion rises, reduce the fan level when the temperature rise is controllable; When the energy-saving heat dissipation period proportion rises, suspend step adjustment; When the heat dissipation energy-consuming period proportion rises, continuously reduce the fan level according to the numerical ratio of the heat dissipation speed to the temperature rise speed.

8. The computer fan step governing control system based on multi-sensor feedback according to claim 1, characterized in that, The fault-tolerant analysis replacement unit's fault-tolerant control method includes: When the sensor monitoring data is abnormal, perform fault-tolerant analysis and replacement of the generated adjustment control instruction; When the sensor monitoring data is occasional data, collect the floating trend of the same type of data at adjacent historical moments, combine the interval period between adjacent historical moments and the current moment to detect whether the influence data is floating; if the influence data is not floating and the floating trend of the same type of data at the next moment is synchronized with the preset actual value, it is determined that it is occasional data, the current period adjustment instruction is not executed, and the data is marked as occasional data in the work log.

Images