Intelligent monitoring system for internal temperature of case

Abstract

The invention relates to the technical field of computer hardware monitoring, in particular to a case internal temperature intelligent monitoring system which comprises a temperature acquisition synchronization module, an environment difference analysis module, a cold control response judgment module, a heat source positioning recognition module and an energy consumption load evaluation module. According to the invention, by performing uniform timestamp sorting and collection on the temperature data at different positions, the time consistency of multi-node temperature information is improved, high-precision dynamic tracking of temperature change is realized, the timeliness and integrity of a time sequence are ensured, and the accuracy of the temperature change is improved. The temperature difference between the nodes and the environment and the time distribution characteristics of the nodes support accurate identification and positioning of thermal anomalies, fan response opportunities are judged according to the peak value and fluctuation rate of the temperature difference, fan numbers and control paths are calibrated, the pertinence of response is enhanced, and the core position of a heat source is positioned through temperature difference space comparison of adjacent nodes. The fan rotating speed and current load change analysis establishes dynamic association between a heat source and energy consumption, and thermal anomaly and energy efficiency trend evaluation is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of computer hardware monitoring, and particularly to an intelligent temperature monitoring system for the interior of a chassis. Background Art

[0002] The technical field of computer hardware monitoring includes technical means for real-time detection and management of the operating states of various components of a computer, mainly covering the monitoring and recording of physical parameters such as temperature, voltage, current, and rotation speed of core components such as the central processing unit, motherboard, power supply, storage device, and cooling system. This technical field realizes comprehensive monitoring and management of the hardware operating state through methods such as sensor acquisition, signal transmission, circuit control, and data processing. In practical applications, computer hardware monitoring can not only provide the reading and display of hardware operating data, but also execute control instructions according to preset logic, including operations such as fan speed regulation, alarm, and forced shutdown, thereby ensuring the safe operation of the device. With the improvement of hardware integration and the increase in the complexity of heat dissipation management, this technical field has also continuously introduced means such as multi-channel detection, distributed temperature control, and remote data reading to meet the refined requirements for hardware state monitoring in different application environments.

[0003] Among them, an intelligent temperature monitoring system for the interior of a chassis refers to setting multiple temperature acquisition nodes at different positions to obtain the temperature information of each area inside the chassis in real time, using an embedded controller to perform arithmetic processing on the acquired data, and then making a logical comparison between the temperature change situation and a preset threshold value to determine the temperature state of each acquisition point. This system uses a thermistor as a temperature sensor, obtains a voltage signal through an analog-to-digital conversion circuit and then transmits it to the controller. The controller executes data conversion and integration according to a multi-point sampling program, judges whether there is an abnormal temperature rise through internal instructions, and outputs a digital signal for display or to drive the execution device to start. The overall monitoring function is realized relying on hardware circuit construction, data acquisition logic programming, and multi-point temperature discrimination strategies.

[0004] The prior art mainly relies on single-point temperature monitoring and static parameter comparison, lacking unified processing of data continuity in the time series, making it difficult to form a complete temperature change trajectory across nodes and regions, resulting in a lag in the response to thermal anomalies. In the multi-node monitoring scenario, there are synchronization deviations in sensor data in the time dimension, and the data integration lacks timing verification, easily leading to misjudgment and missed reports. The prior art mostly uses fixed criteria to judge temperature thresholds and cannot adjust the monitoring sensitivity according to environmental differences or operating states, resulting in a lack of effective identification of sudden or continuous temperature increases. At the level of cold control execution, the fan speed regulation logic relies on preset rules, the response mechanism is single, and there is a lack of a dynamic response strategy based on the change of real-time heat source intensity. Heat source localization is mostly based on manual investigation after temperature threshold alarms or simplified model inference, making it difficult to accurately determine the core position of the heat source, especially with low positioning accuracy in complex thermal interference scenarios. Traditional monitoring means ignore the energy consumption change trend in the thermal anomaly area and cannot provide an association analysis between fan efficiency and current load, limiting the ability of energy consumption optimization and heat dissipation optimization. This easily leads to untimely identification of potential faults, low resource allocation efficiency, and excessive operating load of the heat dissipation system, and may even cause system performance degradation or hardware damage in severe cases. Summary of the Invention

[0005] An object of the present invention is to solve the deficiencies in the prior art and propose an intelligent temperature monitoring system for the interior of a chassis.

[0006] To achieve the above object, the present invention adopts the following technical solutions: An intelligent temperature monitoring system for the interior of a chassis includes: The temperature acquisition synchronization module extracts the real-time temperature values of key areas inside the chassis based on the monitoring data inside the chassis, sorts the timestamps of the sensing device data in sequence, collects the temperature data, and filters out abnormal data with a time interval exceeding the set range of the sampling period to generate a temperature time series for the interior of the chassis. The environmental difference analysis module calls the temperature time series for the interior of the chassis, identifies the differences between the node temperatures and the environmental temperatures, filters out the nodes exceeding the synchronization analysis reference threshold, extracts the difference amplitude and time distribution characteristics, and generates a chassis temperature difference distribution table. The cold control response determination module extracts the peak value of the node temperature difference and the fluctuation rate based on the chassis temperature difference distribution table, detects whether the difference rising rate exceeds the cold control trigger reference, filters out the trigger nodes, and calibrates the corresponding fan numbers and control unit indexes to generate a chassis cold control response trigger list. The heat source localization and identification module calls the chassis cold control response trigger list, extracts the positions of abnormal nodes and the temperature values of adjacent nodes, compares the temperature values of adjacent nodes with the temperature value of the central abnormal node, determines the temperature rise points within the local temperature rise range, and locates the core position of the heat source to obtain a local heat source localization map for the chassis.

[0007] As a further solution of the present invention, the internal temperature time series of the chassis includes the main board temperature data series, the heat distribution curve of the power supply area, the temperature response trajectory of the CPU fan, and the heat change record of the hard disk component. The chassis temperature difference distribution table includes the temperature deviation interval distribution, the difference time period annotation, and the out-of-tolerance node index set. The chassis cold control response trigger list includes the cold control trigger node number, the fan control unit index, and the response intensity level identifier. The local heat source positioning map of the chassis includes the abnormal heat source coordinates, the adjacent temperature rise range annotation, and the local temperature gradient layer.

[0008] As a further solution of the present invention, the temperature acquisition synchronization module includes: The sensing data acquisition sub-module monitors the real-time temperature values of the main board temperature sensor, the power supply area sensor, the CPU radiator fan, and the hard disk component sensor based on the internal monitoring data of the chassis, extracts the timestamps of the sensing devices, sorts the temperature data in ascending order of the timestamps, and generates a temperature time series dataset. The abnormal data screening and analysis sub-module calls the temperature time series dataset, identifies the interval difference between adjacent data timestamps, compares the interval difference with the preset sampling period threshold, eliminates the abnormal data points with the interval difference exceeding the threshold, and obtains an effective temperature dataset. The time series data collection and integration sub-module aligns the main board temperature, power supply temperature, fan temperature, and hard disk temperature data based on the effective temperature dataset according to the timestamps, and integrates them into multi-dimensional temperatures under a unified time axis to generate the internal temperature time series of the chassis.

[0009] As a further solution of the present invention, the environmental difference analysis module includes: The temperature difference node identification sub-module calls the internal temperature time series of the chassis and the environmental temperature series, identifies the temperature difference value of the node at each moment, screens the nodes exceeding the synchronous analysis reference threshold, counts the over-limit frequency and the continuous interval, and generates an out-of-tolerance monitoring node set. The difference intensity calculation sub-module extracts the temperature series of the node within the temperature difference abnormal interval according to the out-of-tolerance monitoring node set, identifies the absolute value of the temperature difference, the change intensity, and the cumulative change amount, calculates the node difference intensity value, and obtains the difference intensity distribution index. The abnormal time period extraction sub-module locates the sudden increase interval of the node difference intensity according to the difference intensity distribution index, counts the continuous period and the time distribution density of the temperature difference abnormality, aggregates the high-frequency abnormal sections of multiple nodes within the same period, and generates the chassis temperature difference distribution table.

[0010] As a further solution of the present invention, the cold control response determination module includes: The difference feature extraction sub-module extracts the temperature difference peak value and the fluctuation rate of the node based on the chassis temperature difference distribution table, and generates a difference feature set. The rate determination sub-module calls the fluctuation rate in the difference feature set, detects whether the difference rising rate exceeds the cold control trigger benchmark, calculates the dynamic response index, linearly compares it with the benchmark value, and generates an over-limit node identification set; The trigger calibration sub-module filters trigger nodes based on the over-limit node identification set, analyzes the mapping relationship between the fan number and the control unit index, and generates a chassis cold control response trigger list.

[0011] As a further solution of the present invention, the heat source positioning and identification module includes: The abnormal node extraction sub-module calls the chassis cold control response trigger list, identifies the coordinates of abnormal nodes and the numbers of adjacent nodes, collects the temperature data of adjacent nodes, compares it with the temperature value of abnormal nodes, and identifies the area where the temperature difference exceeds the abnormal determination threshold to obtain the abnormal adjacent node temperature difference interval value; The local temperature rise identification sub-module calls the abnormal adjacent node temperature difference interval value, filters continuously temperature difference mutation nodes, identifies the temperature difference gradient and the center offset in the section, and determines whether it is concentratedly distributed along the dominant direction to obtain the temperature difference mutation section gradient; The heat source core positioning sub-module analyzes the node space coordinates and the temperature aggregation trend index associated with the temperature difference direction according to the temperature difference mutation section gradient, calculates the heat source center space positioning value, overlays the node density map and the main heat conduction channel, identifies the coordinates of the high-temperature aggregation area, and obtains the chassis local heat source positioning map.

[0012] As a further solution of the present invention, the system further includes an energy consumption load evaluation module: The energy consumption load evaluation module extracts the change amount of the fan speed and the change value of the current load in the heat source area according to the chassis local heat source positioning map, analyzes the percentage increase in speed and the increase amplitude of the current per unit time, and merges the energy consumption level of the node fan and the load change trend to obtain the chassis local energy consumption load trend table; The chassis local energy consumption load trend table includes the fan speed increase amplitude curve, the current load change trend, and the fan energy efficiency ratio analysis result.

[0013] As a further solution of the present invention, the energy consumption load evaluation module includes: The heat source data identification sub-module extracts the change data of the fan speed and the current load in the heat source area according to the chassis local heat source positioning map, intercepts continuous time series in seconds, calculates the percentage increase in speed and the increase amplitude of the current at adjacent time points, analyzes the relationship between the time stamp and the increment value, and generates a heat source increment data set; The speed-current correlation sub-module calls the heat source increment data set, divides the data segments according to the time window, performs linear regression analysis on the percentage increase in speed sequence and the current increase amplitude sequence in each window, extracts the regression slope as the correlation intensity value, and generates a correlation coefficient matrix; Based on the correlation coefficient matrix, the energy consumption trend merging sub-module combines the percentage of rotational speed increment per unit time with the correlation intensity value, superimposes the current increment amplitude, and integrates the time series data according to the nodes in the heat source area to obtain the local energy consumption load trend table of the chassis.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In the present invention, by sorting and aggregating the data collected by temperature sensors at different positions with a unified time stamp, the consistency of multi-node temperature information in the time dimension is improved, enabling the system to achieve high-precision dynamic tracking of temperature changes. The data with abnormal sampling periods are screened out, effectively ensuring the timeliness and integrity of the temperature time series and providing a stable basis for subsequent analysis. In the environmental temperature difference analysis, by extracting the temperature differences between nodes and the environment and their time distribution characteristics, accurate identification and positioning of thermal anomalies are achieved, enabling the temperature control strategy to be dynamically adjusted according to the difference amplitude and duration characteristics. The peak value and fluctuation rate of the temperature difference are used to determine whether to trigger the fan control response, and the fan number and control path are synchronously calibrated, enhancing the pertinence and timeliness of the thermal management response. During the abnormal heat source positioning process, by making a spatial comparison of the temperature differences between abnormal nodes and adjacent areas, a heat source identification mechanism centered on the local temperature rise range is constructed, effectively improving the accuracy of heat source tracing. Finally, in the energy consumption load assessment, by introducing the analysis of the changes in fan speed and current load, the heat source intensity and energy consumption are dynamically correlated, realizing a comprehensive assessment from thermal anomalies to energy efficiency trends. This processing flow is centered on refined data sorting and cleaning, high-dimensional temperature difference judgment, local heat source modeling, and energy consumption response analysis, constructing a data-driven, response-closed-loop, high-resolution temperature control monitoring system, greatly enhancing the intelligence, real-time performance, and energy-saving efficiency of heat dissipation control. Description of the Drawings

[0015] Figure 1 is the system flow chart of the present invention; Figure 2 is the flow chart of the temperature acquisition synchronization module in the present invention; Figure 3 is the flow chart of the environmental difference analysis module in the present invention; Figure 4 is the flow chart of the cooling control response determination module in the present invention; Figure 5 is the flow chart of the heat source positioning and identification module in the present invention; Figure 6 is the flow chart of the energy consumption load assessment module in the present invention. Detailed Embodiments

[0016] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are based on the orientation or positional relationships shown in the accompanying drawings. These are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more unless otherwise specifically defined.

[0018] Please refer to Figure 1 , an intelligent monitoring system for the internal temperature of a chassis includes: The temperature acquisition synchronization module extracts the real-time temperature values of the motherboard temperature sensor, power supply area sensor, CPU radiator fan, and hard disk component sensor based on the monitoring data inside the chassis, sorts the timestamps of the sensing device data in sequence, aggregates the temperature data, and filters out abnormal data with a time interval exceeding the set range of the sampling period to generate a time series of the internal temperature of the chassis. The environmental difference analysis module calls the time series of the internal temperature of the chassis, identifies the differences between the node temperatures and the environmental temperature, filters out the nodes exceeding the synchronization analysis reference threshold, extracts the difference amplitude and time distribution characteristics, and generates a chassis temperature difference distribution table. The cold control response determination module extracts the peak value of the node temperature difference and the fluctuation rate based on the chassis temperature difference distribution table, detects whether the difference rising rate exceeds the cold control trigger reference, filters out the trigger nodes, and calibrates the corresponding fan numbers and control unit indices to generate a chassis cold control response trigger list. The heat source location and identification module calls the chassis cold control response trigger list, extracts the abnormal node positions and the temperature values of adjacent nodes, compares the temperature values of adjacent nodes with the temperature value of the central abnormal node, determines the temperature rise points within the local temperature rise range, and locates the core position of the heat source to obtain a local heat source location map of the chassis. The energy consumption and load assessment module extracts the change amount of the fan speed and the change value of the current load within the heat source area according to the local heat source location map of the chassis, analyzes the percentage increase in speed and the increase amplitude of the current per unit time, and merges the energy consumption levels of the node fans and the load change trends to obtain a local energy consumption and load trend table of the chassis.

[0019] The internal chassis temperature time series includes the mainboard temperature data series, the power supply area heat distribution curve, the CPU fan temperature response trajectory, and the hard disk component heat change record. The chassis temperature difference distribution table includes the temperature deviation interval distribution, the difference time period annotation, and the out-of-tolerance node index set. The chassis cold control response trigger list includes the cold control trigger node number, the fan control unit index, and the response intensity level identifier. The chassis local heat source location map includes the abnormal heat source coordinates, the adjacent temperature rise range annotation, and the local temperature gradient layer. The chassis local energy consumption load trend table includes the fan speed increase curve, the current load change trend, and the fan energy efficiency ratio analysis result.

[0020] Please refer to Figure 2 , the temperature acquisition synchronization module includes: The sensing data acquisition sub-module monitors the real-time temperature values of the mainboard temperature sensor, the power supply area sensor, the CPU radiator fan, and the hard disk component sensor based on the internal chassis monitoring data, extracts the timestamps of the sensing devices, sorts the temperature data in ascending order of the timestamps, and generates a temperature time series data set. Collect the data of each sensor inside the chassis. By reading the real-time temperature values of the mainboard temperature sensor, the power supply area sensor, the CPU radiator fan, and the hard disk component sensor, obtain the original temperature data of each sensor. For the read timestamp of each sensor, first perform precise parsing on the timestamp to ensure the accuracy and timeliness of the data. According to the collected timestamp, sort the sensor data to ensure that the temperature data arranged in chronological order can reflect the actual chassis temperature change process. For example, assume that at a certain moment, the temperature recorded by the mainboard temperature sensor is 45°C, the temperature recorded by the power supply area temperature sensor is 50°C, the temperature recorded by the CPU radiator fan is 48°C, and the temperature recorded by the hard disk component temperature sensor is 42°C, and the timestamps are T1, T2, T3, T4 respectively. During the sorting process, all temperature data will be arranged in ascending order according to the T1, T2, T3, T4 timestamps to ensure that all data is displayed in chronological order. By merging the above sorted data, generate a temperature time series data set to provide a basis for subsequent data processing.

[0021] The abnormal data screening and analysis sub-module calls the temperature time series data set, identifies the interval difference between adjacent data timestamps, compares the interval difference with the preset sampling period threshold, eliminates the abnormal data points whose interval difference exceeds the threshold, and obtains a valid temperature data set. Call the data in the temperature time-series dataset and check the time interval between the timestamps of every two adjacent data points. Assume the preset sampling period is 2 seconds. That is to say, it is expected that the time interval between adjacent data points should not be greater than 2 seconds. If the time difference between two data points exceeds 2 seconds, then this data point is considered abnormal and should be removed. For example, assume there are temperature data for two data points between timestamps 10:00:01 and 10:00:05, and the time interval is 4 seconds, which obviously exceeds the preset 2-second sampling period. Then one or both of these two data points will be determined as abnormal data points and need to be removed from the temperature dataset. Through programming logic, filter out all temperature data points with interval differences exceeding the threshold, and only retain the valid temperature dataset with time intervals meeting the expectations, ensuring that the time-series data in the temperature dataset is continuous and accurate and can reflect the real temperature changes.

[0022] The time-series data collection sub-module aligns the motherboard temperature, power supply temperature, fan temperature, and hard disk temperature data based on the valid temperature dataset according to the timestamps, and integrates them into multi-dimensional temperatures under a unified timeline to generate the internal chassis temperature time series. Extract the datasets of each sensor from the valid temperature dataset, mainly including motherboard temperature, power supply temperature, fan temperature, and hard disk temperature data. The data is aligned according to the unified timeline based on the timestamps respectively. If a certain sensor has no data at a certain time point (for example, a certain sensor fails to collect data due to a fault or data loss), interpolation method will be used to fill the data at this time point, or the value of the previous or next data point will be used for substitution to ensure the integrity of the data. Assume that at a certain moment, the motherboard temperature is 32 °C, the power supply temperature is 35 °C, the fan temperature is 30 °C, and the hard disk temperature is 28 °C. If a certain sensor data is missing at this moment, interpolation is performed based on the data before and after to generate a reasonable estimated value. The process will ensure that the data of all sensors are arranged in the same time order to generate a unified temperature time-series dataset. On this basis, finally obtain the internal chassis temperature time series, which will combine the multi-dimensional temperature data of the motherboard, power supply, fan, and hard disk to provide data support for the monitoring and analysis of the internal chassis temperature.

[0023] Please refer to Figure 3 , the environmental difference analysis module includes: The temperature difference node identification sub-module calls the internal chassis temperature time series and the environmental temperature series, identifies the temperature difference value of the node at each moment, screens out the nodes exceeding the synchronous analysis benchmark threshold, counts the overlimit frequency and the continuous interval, and generates the over-difference monitoring node set. The data is collected through real-time monitoring by sensors. The temperature data of each node at different time points is obtained through dedicated monitoring equipment, such as temperature sensors. The ambient temperature is measured outside the chassis by the same equipment to ensure the real-time and accuracy of the data. The monitoring records the data every certain time period, such as every 10 minutes, to form a time series. The data is stored in the monitoring database for subsequent processing. The temperature difference of each node at each time is calculated by extracting the corresponding temperature record from the database and subtracting the internal node temperature from the ambient temperature at each time point. For example, if the node temperature is 35°C and the ambient temperature is 30°C, then the temperature difference is 5°C. The steps of screening nodes that exceed the synchronous analysis benchmark threshold are based on the preset threshold. If the temperature difference threshold is set to 4°C, all nodes with a temperature difference greater than 4°C will be marked and recorded. The frequency and duration of exceeding the limit are counted by further analyzing the filtered data, recording the number of times each node exceeds the threshold and the length of time each time the threshold is exceeded, and generating a set of out-of-tolerance monitoring nodes. This process not only helps identify nodes with abnormal temperatures, but also evaluates the temperature stability of the nodes, providing a basis for subsequent temperature control management.

[0024] The difference intensity calculation submodule extracts the temperature sequence of the node within the temperature difference abnormal range according to the out-of-tolerance monitoring node set, identifies the absolute value of the temperature difference, the change intensity and the cumulative change, and uses the formula: ; Calculate the node difference strength value and obtain the difference strength distribution index; in, is the node difference strength value, For Node At the moment The temperature, For the corresponding time The ambient temperature, For Node The average temperature, is the node temperature change rate, is the total duration of the moment; The node difference intensity value indicates the degree of temperature fluctuation of a single node inside the chassis relative to the ambient temperature within a specific time period, reflecting the cumulative amplitude of the temperature anomaly, the severity of the fluctuation, and the composite level of the temperature change rate of the node. The larger the value, the more drastic the temperature change of the node, the more unstable the fluctuation, and the higher the degree of deviation from the ambient temperature. Therefore, it can be used to identify abnormal temperature control hotspots and assist in the subsequent chassis thermal management strategy formulation and key area diagnosis. The data in the out-of-tolerance monitoring node set is classified and processed to extract the temperature sequence of each node within the temperature difference abnormal range. For example, the temperature sequence recorded by node A in a certain period of time is , the environmental temperature sequence for the corresponding time period is , and the absolute value of the temperature difference is calculated by taking the absolute value of the difference between the node temperature and the environmental temperature at the corresponding moment. For example, the absolute value of the temperature difference at the first time point is ; For the calculation of the change intensity, considering the need to evaluate the rapid temperature change, the change rate of the node temperature can be calculated. For example, if the node temperature changes from 32°C to 35°C in 60 minutes, its change intensity can be expressed as ; The cumulative change amount is the accumulation of the absolute values of the temperature differences over a period of time. In the above example, if the sequence of absolute values of the temperature differences at each time point is , then the cumulative change amount is ; Now, the following formula is used to comprehensively consider the absolute value of the temperature difference, the change intensity, and the cumulative change amount to provide a quantified difference intensity value. To ensure the consistency and comparability of the data in the formula, the following operations are performed: Absolute value of the temperature difference: The sequence of absolute values of the temperature differences is directly used for calculation; Average node temperature : Calculate the average value of the node temperature sequence ; Rate of change of the temperature difference : Convert the rate to the reciprocal of hours, that is, convert 0.05°C / minute to ; Now, substitute specific values to calculate the node difference intensity value : Cumulative absolute value of the temperature difference: ; Standard deviation of the node temperature difference: ; Total time , total rate of change of the temperature difference °C / hour; Substitute the values into the formula for calculation: ; Therefore, is the comprehensive difference intensity value of the node, which characterizes the intensity and volatility of the temperature difference of the node during the monitoring period. In this way, technicians can quickly evaluate and identify the hot spots that need special attention and perform corresponding temperature control optimization.

[0025] The abnormal period extraction sub-module locates the sudden increase interval of the node difference intensity according to the difference intensity distribution index, counts the duration and time distribution density of the temperature difference abnormality, aggregates the high-frequency abnormal sections of multiple nodes within the same period, and generates a chassis temperature difference distribution table; Based on the data in the differential intensity distribution index, identify the time periods and nodes where the temperature difference intensity increases sharply within a short time, and count the duration and time distribution density of abnormal temperature differences. This is to perform a time analysis on the identified data in the previous step, calculating the total duration of abnormal temperature differences for each abnormal node and the distribution of time throughout the monitoring period. For example, if a certain node often shows abnormal temperature differences during the midnight period, which is related to the decrease in ambient temperature, aggregate the high-frequency abnormal sections of multiple nodes within the same cycle. This is to merge the data of multiple nodes with similar abnormal patterns, analyze whether there are common temperature control problems or environmental factor influences among the nodes, and generate a chassis temperature difference distribution table. This table shows the abnormal temperature difference distributions of all abnormal nodes at different time periods, providing important data for the optimization of chassis temperature control.

[0026] Please refer to Figure 4 , the cold control response determination module includes: The differential feature extraction sub-module extracts the peak value and fluctuation rate of node temperature differences based on the chassis temperature difference distribution table, generating a differential feature set; Collect the temperatures of each node inside the chassis. The nodes are different positions inside the chassis, such as beside the fan, above the motherboard, and in the hard disk area, etc. Obtain a series of temperature values from the data collected by temperature sensors at different time points. By comparing the temperature values, calculate the temperature differences between each node, and record the maximum temperature difference peak value of each node. Based on time series analysis, calculate the fluctuation rate of each node's temperature, that is, the rate of temperature change, measured by the rate of change of temperature over time. If the node temperature changes rapidly, its fluctuation rate is high. By extracting the temperature difference peak value and fluctuation rate, a differential feature set is obtained for subsequent judgment and analysis. For example, if the temperature peak difference at a certain node exceeds 5°C, and the temperature fluctuation rate of this node exceeds 0.2°C / s, then the temperature difference feature of this node will be marked as exceeding the normal fluctuation, serving as the basis for subsequent calculations and judgments.

[0027] The rate determination sub-module calls the fluctuation rate in the differential feature set to detect whether the differential increase rate exceeds the cold control trigger benchmark, using the formula: ; Calculate the dynamic response index, perform a linear comparison with the benchmark value, and generate an over-limit node identification set; Among them, represents the dynamic response index, represents the peak value of node temperature difference, represents the fluctuation rate, represents the cold control trigger benchmark; The dynamic response index is a quantitative indicator used to measure the relationship between the temperature change rate of the device node and the response of the cooling control system. It is calculated by combining the temperature difference peak value (ΔP) and the temperature fluctuation rate (V). It reflects the severity of the node temperature change and the urgency of its response to the cooling control system. The temperature difference peak value ΔP indicates the maximum temperature fluctuation of the node, and the fluctuation rate V indicates the speed of temperature change. The two together determine whether the cooling control system needs to be started. If the dynamic response index R exceeds the cooling control trigger reference value B, it means that the temperature change of the node is too drastic and the cooling control system should intervene in time to cool down. This means that the temperature change is within an acceptable range and no additional intervention is required. The calculation of this index helps to monitor the device temperature in real time and ensure that the system can maintain stable operation in a high temperature environment, thereby avoiding failures caused by equipment overheating. First, we need to extract the fluctuation rate (V) and temperature difference peak (ΔP) of each node from the difference feature set, and substitute the values ​​into the formula for calculation; ΔP (peak temperature difference): ΔP is the peak temperature difference of each node. It is calculated from the data collected by the temperature sensor. The temperature value of each node is obtained from multiple time points, and then the maximum difference between the temperature values ​​is calculated. For example, the temperature range of a node is from 25°C to 31°C, and the maximum temperature difference ΔP is 6°C. This is the quantitative result of the node temperature difference. V (fluctuation rate): V is the rate at which the node temperature changes over time, measured by the slope of the temperature change. Suppose the temperature of a node rises from 26°C to 27°C in 1 second. The fluctuation rate V is calculated as: (27°C-26°C) / 1 second = 1°C / s. The temperature fluctuation rate is the ratio of time to temperature change, and the unit is ℃ / s; B (cold control trigger benchmark): B is the cold control trigger benchmark, which indicates the threshold of temperature change. It is generally set according to the design of the equipment or the working environment conditions of the system. B can be obtained through experimental data or original data and will be set as the optimal cooling working condition when the equipment is designed. Assume that when the cold control system is designed, the setting value of B is 3°C, which means that when the node temperature difference exceeds 3°C, the cooling system will start; In order to make the parameters effectively calculated in the formula, it is necessary to ensure that the dimensions of all parameters are unified. First, the temperature difference ΔP and the fluctuation rate V need to be converted to the same unit. The units of the temperature difference ΔP and the fluctuation rate V are consistent. ΔP is ℃, and V is ℃ / s. Therefore, they can be directly substituted into the formula without additional conversion. In order to avoid inconsistent units affecting the calculation, normalization is performed to convert the units of V and ΔP into unified standard units. For example, by using relative rate of change or ratio calculation, the temperature difference and the fluctuation rate are processed as dimensionless data. Suppose the values extracted from the set of differential features are as follows: ΔP = 6 °C (the peak value of the maximum temperature difference at a certain node), V = 0.4 °C / s (the fluctuation rate at a certain node), B = 3 °C (the cold control trigger reference value); Substitute the values: ; The obtained dynamic response index R is approximately 0.208. This result needs to be compared with the preset cold control trigger reference value. Suppose the set cold control reference value is 0.25. Obviously, 0.208 is lower than 0.25. Therefore, the cold control reaction of this node has not been triggered. This result indicates that the temperature change rate of the current node is low, and its dynamic response index does not exceed the trigger threshold of the cold control system. Therefore, in the current environment, the cooling requirement of this node is not high, and there is no need to start the cold control system. Through the calculation process, it can help the system to monitor and evaluate the temperature change of each node in real time and decide whether to start the fan or cooling equipment.

[0028] The trigger calibration sub-module filters the trigger nodes based on the set of over-limit node identifiers, analyzes the mapping relationship between the fan numbers and the control unit indexes, and generates a chassis cold control response trigger list; According to the set of over-limit node identifiers, filter out those nodes whose responses exceed the reference value. The temperature change of the nodes is too drastic, which affects the stability of the equipment and needs to be adjusted through the cooling system. Use the mapping relationship between the fan numbers and the indexes of the control unit to associate the over-limit nodes with specific cooling system units. Suppose the over-limit identifier of node A is 1, indicating that this node needs cooling. Then, through the index mapping relationship, the corresponding fan number or cooling unit of this node can be found. Suppose the fan number is 5, then the fan number 5 will be added to the trigger list. By continuously filtering and index mapping all the nodes, a chassis cold control response trigger list is generated, which lists all the nodes that need to enable cooling and their corresponding cooling unit numbers. The cooling system starts the corresponding fans or cooling modules according to this list to ensure the stable operation of the equipment in a high-temperature environment.

[0029] Please refer to Figure 5 , the heat source location and identification module includes: The abnormal node extraction sub-module calls the chassis cold control response trigger list, identifies the coordinates of the abnormal nodes and the numbers of adjacent nodes, collects the temperature data of the adjacent nodes, compares it with the temperature values of the abnormal nodes, identifies the area where the temperature difference exceeds the abnormal determination threshold, and obtains the abnormal adjacent node temperature difference interval value; Obtain the coordinates and temperature data of abnormal nodes and neighboring nodes from within the chassis. This is a common operation in the heat dissipation of high-performance computer servers or large data centers. In the example, assume that when a server is performing large-scale data processing, it is found that the temperature of a certain CPU core is abnormal. This node is automatically marked as an abnormal node, and its coordinates are obtained, as well as the temperature data of the neighboring nodes. The data is transmitted to the central control in real time through the sensor network. The central control system compares the data with the original data and determines which neighboring nodes have a temperature difference exceeding the normal range by setting a temperature difference threshold (for example, setting the threshold to 5°C), thereby generating a temperature difference range value for the nodes. This range value is the difference between the abnormal temperature value and the average temperature value of the normal nodes. For example, if the temperature of the abnormal node is 80°C and the average temperature of the normal nodes is 72°C, then the temperature difference range value is 8°C, exceeding the set threshold. The process does not involve complex calculations or algorithms, but directly compares and judges through the set operation of sensor data to ensure the real-time and accuracy of the operation, generates the temperature difference range value of abnormal neighboring nodes, and provides key data to support the further formulation of fault diagnosis and prevention measures.

[0030] The local temperature rise identification sub-module calls the temperature difference range value of abnormal neighboring nodes, screens for continuously temperature difference mutation nodes, identifies the temperature difference gradient and the center offset within the section, and determines whether they are concentrated in the dominant direction to obtain the temperature difference mutation section gradient. Sort the values and identify continuously temperature difference mutation nodes that exceed the set threshold. For example, in a specific case, if within a region of the chassis layout, the temperature difference mutation values of three consecutive nodes are 6°C, 7°C, and 8°C respectively, and the values all exceed the set distribution balance threshold of 5°C, then the section is marked as a key monitoring area. Calculate the temperature difference distribution gradient value of this section, that is, the temperature difference change rate of consecutive mutation points, and the offset value relative to the central node. The calculation uses simple mathematical formulas. For example, the gradient value calculation formula is (temperature difference value n - temperature difference value n - 1) / node spacing, and the offset is calculated based on the physical position coordinates of the nodes. Through basic calculations, it can be determined whether the temperature difference mutation is concentrated in one direction, thereby diagnosing the potential heat source direction and generating the temperature difference mutation section gradient, providing a data basis for further accurately locating the heat source.

[0031] The heat source core positioning sub-module, based on the temperature difference mutation section gradient, analyzes the node space coordinates and temperature aggregation trend indicators associated with the temperature difference direction, using the formula: ; Calculate the heat source center space positioning value, overlay the node density map and the main heat conduction channel, identify the coordinates of the high-temperature aggregation area, and obtain the local heat source positioning map of the chassis. Among them, represents the heat source center space positioning value, represents the Temperature difference gradient of the temperature difference node Denote the Horizontal offset value of the temperature difference node Denote the Vertical offset value of the temperature difference node Represent the Temperature trend weight of the spatial distribution node For the Heat gradient value of the spatial distribution node Is the total number of temperature difference nodes Is the total number of spatial distribution nodes; The spatial positioning value of the heat source center is a key indicator that quantifies the spatial position and heat concentration degree of the heat source in the chassis by comprehensively considering multiple factors such as the temperature difference gradient between nodes, spatial offset, and temperature concentration trend. The calculation of this value is based on the temperature difference change (i.e., temperature gradient) of each node, the spatial position of the node (i.e., horizontal and vertical offsets relative to the reference point), and the temperature aggregation degree weight of each node, and a comprehensive result is obtained through weighted summation of the parameters. Its value represents the core position of the heat source in space. The higher the value, the more significant the temperature change in this area, the stronger the heat source aggregation effect, which means that this position is a potential hot spot area; According to the temperature difference mutation section gradient, extract the temperature difference gradient values and their corresponding spatial coordinate information of each node in this section, and construct a spatial heat source positioning structure with the heat concentration trend as the dominant factor. During the execution process, it is necessary to first obtain the temperature difference gradient values of each node , and the acquisition method is: take the difference between the temperature difference values of two adjacent nodes in the mutation section and divide it by the spatial distance between the two nodes. For example, if the temperatures of nodes A and B are 84.3 °C and 78.0 °C respectively, and the distance is 1.5 cm, then the corresponding temperature difference gradient is , then set the origin according to the center position of the node, and calculate the offset value of the coordinates of each node relative to this origin , for example, the horizontal offset of node B is 2 cm and the vertical offset is 0 cm, obtaining , , then obtain the temperature trend index weight , and its setting basis is the temperature proportion of the node temperature in its local area. The calculation method is the ratio of the node temperature to the average temperature of all nodes in this local area. After normalization processing, the weight value floats within the range of [0, 1]. Assuming that the local area average is 76.5 °C and the temperature of node B is 84.3 °C, then the initial value is , and after normalization, take , at the same time, the heat conduction gradient Is obtained by dividing the temperature difference value from this node to the adjacent lower-level node by the heat conduction distance. Assuming that the temperature of the next node is 72.0 °C and the distance is 1.5 cm, then After unifying the dimensions of all parameters, they are normalized; Substitute into the formula: ; By simultaneously introducing the dual weighted fusion of the temperature difference gradient and the heat conduction path, the spatial positioning accuracy of the heat source core area is further quantitatively supported. The results show that the heat source is concentrated at the superposition coordinates of the node density structure, and its position and numerical characteristics can be used for subsequent heat dissipation scheduling or heat control path reconstruction.

[0032] Please refer to Figure 6 , the energy consumption load evaluation module includes: The heat source data identification sub-module extracts the change data of the fan speed and current load in the heat source area according to the local heat source positioning map of the chassis, intercepts continuous time series in seconds, calculates the percentage of speed increment and the amplitude of current increment at adjacent time points, analyzes the relationship between the timestamp and the increment value, and generates a heat source increment data set; Identify the location of the heat source area, and then extract relevant data through multiple physical parameters such as temperature, wind speed, fan speed, and current. Among them, the change data of fan speed and current load are the key to heat source increment analysis. During the data extraction process, the data within each second will be intercepted to form a continuous time series. In this time series, by calculating the percentage of fan speed increment and the amplitude of current increment at two adjacent time points, the relationship between the timestamp and the increment can be analyzed, and further a heat source increment data set related to the heat source area can be obtained. The increment data set will be used as an important basis for subsequent analysis. By summarizing the increment data in different heat source areas, the analysis of the heat source change trend of the entire system can be finally realized.

[0033] The rotation speed-current correlation sub-module calls the heat source increment data set, divides the data segments according to the time window, performs linear regression analysis on the rotation speed increment percentage sequence and the current increase amplitude sequence within each window, extracts the regression slope as the correlation intensity value, and generates a correlation coefficient matrix; Call the heat source increment data set, divide the data by time window, and perform statistics per second to form small data segments. For each data segment, a linear regression analysis will be performed on the fan speed increment percentage sequence and the current increase sequence. By fitting the regression slope, the correlation strength value is obtained. When specifically executing, first, within each time window, perform corresponding matching on the fan speed increment percentage sequence and the current increase sequence. Suppose within a certain time window, the fan speed increases from 800 rpm to 850 rpm, with an increment of 6.25%, and the current increases from 1.5 A to 1.6 A, with an increase of 6.67%. Then the linear regression analysis within this window will calculate the regression slope between these two parameters, thereby obtaining the correlation strength within this time period. For the data of all time windows, a correlation coefficient matrix is generated to represent the correlation degree between the fan speed and the current within different time periods, thus providing data support for energy consumption analysis.

[0034] Based on the correlation coefficient matrix, the energy consumption trend merging sub-module combines the rotational speed increment percentage per unit time with the correlation strength value, superimposes the current increment amplitude, and integrates the time series data according to the heat source area nodes to obtain the chassis local energy consumption load trend table; Combined with the correlation coefficient matrix, integrate the fan speed increment percentage per unit time and the correlation strength value, and superimpose the current increment amplitude to generate the energy consumption load trend data within the heat source area. When specifically executing, first, according to the correlation strength value calculated in the correlation coefficient matrix, perform a weighted calculation on the fan speed increment percentage and the correlation strength value within each time window to obtain a weighted energy consumption load data. Then superimpose the current increment amplitude on the weighted rotational speed increment value to obtain the final energy consumption trend data. Within the heat source area, integrate according to each node, and inductively summarize the energy consumption load trend for each time period one by one to obtain the chassis local energy consumption load trend table, which shows the energy consumption trend of each heat source area over time and provides a basis for subsequent optimization of the system energy consumption.

[0035] The above is only a preferred embodiment of the present invention, and it is not intended to limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An intelligent monitoring system for the internal temperature of a chassis, characterized in that, The system includes: Based on the monitoring data inside the chassis, the temperature acquisition synchronization module extracts the real-time temperature values of key areas inside the chassis, sorts the timestamps of the sensing device data in sequence, aggregates the temperature data, and filters out abnormal data with time intervals exceeding the set range of the sampling period, generating a temperature time series inside the chassis; The environmental difference analysis module calls the temperature time series inside the chassis, identifies the differences between the node temperature and the environmental temperature, filters out the nodes exceeding the synchronization analysis benchmark threshold, extracts the difference amplitude and time distribution characteristics, and generates a chassis temperature difference distribution table; Based on the chassis temperature difference distribution table, the cold control response determination module extracts the peak value of the node temperature difference and the fluctuation rate, detects whether the difference rising rate exceeds the cold control trigger benchmark, filters out the trigger nodes, and calibrates the corresponding fan numbers and control unit indexes, generating a chassis cold control response trigger list; The heat source location and identification module calls the chassis cold control response trigger list, extracts the abnormal node positions and the temperature values of adjacent nodes, compares the temperature values of adjacent nodes with the temperature value of the central abnormal node, determines the temperature rise points within the local temperature rise range, locates the core position of the heat source, and obtains a chassis local heat source location map.

2. The intelligent internal temperature monitoring system for a chassis according to claim 1, characterized in that, The temperature time series inside the chassis includes the motherboard temperature data series, the heat distribution curve of the power supply area, the temperature response trajectory of the CPU fan, and the heat change record of the hard disk component. The chassis temperature difference distribution table includes the temperature deviation interval distribution, the difference time period annotation, and the out-of-tolerance node index set. The chassis cold control response trigger list includes the cold control trigger node numbers, the fan control unit indexes, and the response intensity level identifiers. The chassis local heat source location map includes the abnormal heat source coordinates, the adjacent temperature rise range annotation, and the local temperature gradient layer.

3. The intelligent monitoring system for the internal temperature of the chassis according to claim 1, characterized in that The temperature acquisition synchronization module includes: Based on the monitoring data inside the chassis, the sensing data acquisition sub-module monitors the real-time temperature values of the motherboard temperature sensor, the power supply area sensor, the CPU radiator fan, and the hard disk component sensor, extracts the timestamps of the sensing devices, and sorts the temperature data in ascending order of the timestamps, generating a temperature time series dataset; The abnormal data screening and analysis sub-module calls the temperature time series dataset, identifies the interval difference between adjacent data timestamps, compares the interval difference with the preset sampling period threshold, and eliminates the abnormal data points with interval differences exceeding the threshold, obtaining a valid temperature dataset; Based on the valid temperature dataset, the time series data aggregation sub-module aligns the motherboard temperature, power supply temperature, fan temperature, and hard disk temperature data according to the timestamps, and integrates them into multi-dimensional temperatures under a unified time axis, generating a temperature time series inside the chassis.

4. The intelligent monitoring system for the internal temperature of the chassis according to claim 3, characterized in that, The environmental difference analysis module includes: The temperature difference node identification sub-module calls the temperature time series inside the chassis and the environmental temperature series, identifies the temperature difference value of the node at each moment, filters out the nodes exceeding the synchronization analysis benchmark threshold, counts the over-limit frequency and the continuous interval, and generates an out-of-tolerance monitoring node set; Based on the out-of-tolerance monitoring node set, the difference intensity calculation sub-module extracts the temperature series of the node within the temperature difference abnormal interval, identifies the absolute value of the temperature difference, the change intensity, and the cumulative change amount, calculates the node difference intensity value, and obtains the difference intensity distribution index; The abnormal period extraction sub-module locates the interval with a sudden increase in the node difference intensity according to the difference intensity distribution index, counts the duration and time distribution density of the temperature difference abnormality, aggregates the high-frequency abnormal sections of multiple nodes within the same period, and generates a chassis temperature difference distribution table.

5. The intelligent internal temperature monitoring system for a chassis according to claim 4, wherein, The cooling control response determination module includes: The difference feature extraction sub-module extracts the peak value and fluctuation rate of the node temperature difference based on the chassis temperature difference distribution table, and generates a difference feature set. The rate determination sub-module calls the fluctuation rate in the difference feature set, detects whether the difference rising rate exceeds the cooling control trigger benchmark, calculates the dynamic response index, linearly compares it with the benchmark value, and generates an over-limit node identification set. The trigger calibration sub-module filters the trigger nodes based on the over-limit node identification set, analyzes the mapping relationship between the fan number and the control unit index, and generates a chassis cooling control response trigger list.

6. The intelligent monitoring system for the internal temperature of the chassis according to claim 5, characterized in that, The heat source location and identification module includes: The abnormal node extraction sub-module calls the chassis cooling control response trigger list, identifies the coordinates of the abnormal nodes and the numbers of adjacent nodes, collects the temperature data of the adjacent nodes, compares it with the temperature values of the abnormal nodes, and identifies the area where the temperature difference exceeds the abnormal determination threshold to obtain the temperature difference interval value of the abnormal adjacent nodes. The local temperature rise identification sub-module calls the temperature difference interval value of the abnormal adjacent nodes, filters the nodes with continuous temperature difference mutations, identifies the temperature difference gradient and the center offset within the section, and determines whether it is concentrated along the dominant direction to obtain the temperature difference mutation section gradient. The heat source core location sub-module analyzes the node spatial coordinates and temperature aggregation trend index associated with the temperature difference direction according to the temperature difference mutation section gradient, calculates the spatial location value of the heat source center, overlays the node density map and the main heat conduction channel, and identifies the coordinates of the high-temperature aggregation area to obtain the chassis local heat source location map.

7. The intelligent internal temperature monitoring system for the chassis according to claim 1, characterized in that The system also includes an energy consumption load assessment module: The energy consumption load assessment module extracts the change amount of the fan speed and the change value of the current load within the heat source area according to the chassis local heat source location map, analyzes the percentage increase in speed and the increase amplitude of the current per unit time, and merges the energy consumption level and load change trend of the node fans to obtain a chassis local energy consumption load trend table. The chassis local energy consumption load trend table includes the fan speed increase amplitude curve, the current load change trend, and the fan energy efficiency ratio analysis result.

8. The intelligent monitoring system for the internal temperature of the chassis according to claim 7, characterized in that The energy consumption load assessment module includes: The heat source data identification sub-module extracts the change data of the fan speed and the current load within the heat source area according to the chassis local heat source location map, intercepts continuous time series in seconds, calculates the percentage increase in speed and the increase amplitude of the current between adjacent time points, analyzes the relationship between the time stamp and the increment value, and generates a heat source increment data set. The speed-current correlation sub-module calls the heat source increment data set, divides the data segments according to the time window, performs a linear regression analysis on the percentage increase in speed sequence and the current increase amplitude sequence within each window, extracts the regression slope as the correlation intensity value, and generates a correlation coefficient matrix. The energy consumption trend merging sub-module combines the percentage increase in speed per unit time and the correlation intensity value based on the correlation coefficient matrix, superimposes the current increase amplitude, and integrates the time series data according to the nodes in the heat source area to obtain a chassis local energy consumption load trend table.

Images