A heating ventilation system operation data visual processing system

Abstract

The present application relates to the technical field of data visualization, in particular to a heating system operation data visualization processing system, the system comprises a heat flux feature extraction module, a direction gradient difference calculation module, a fluctuation amplitude analysis module, a connection attribute regulation module and a dynamic trajectory optimization module.In the present application, through real-time filtering of heat flux density and direction vector calculation, the heat transfer path between multiple nodes can be dynamically associated and mapped, ensuring the continuity of heat flow changes and avoiding the delay and distortion of traditional static graphics when updating data. Using time series smoothing processing, the heat flow trend and periodic changes are accurately captured, and the fluctuation amplitude and direction can be intuitively presented through color and line width changes, showing the heat flow intensity and direction changes. This method enhances the time sequence perception of thermal environment data, improves trend identification and dynamic response capability, and clearly presents the changes in energy transfer process, effectively solving the shortcomings of traditional methods in real-time and spatial relationship expression.

Description

Technical Field

[0001] This invention relates to the field of data visualization technology, and in particular to a data visualization processing system for HVAC system operation. Background Technology

[0002] The field of data visualization technology involves the comprehensive processing of multi-source data collection, organization, analysis, and graphical presentation. Its core components include data acquisition, data cleaning, data classification and correlation analysis, visualization chart generation, and dynamic interactive display. This technology aims to express complex data in an intuitive form through graphical means, enabling users to quickly understand data characteristics and patterns of change. It is widely used in scenarios such as energy management, industrial monitoring, and building operation management.

[0003] Traditional HVAC system operation data visualization processing systems refer to systems that collect data such as temperature, humidity, air volume, water flow, energy consumption, and equipment operating status generated during the operation of building HVAC systems. This data is then acquired through data collection terminals, centrally processed by a data aggregation platform, and displayed using preset chart templates or fixed graphical interfaces. Traditional methods rely on manually setting data parameters and outputting operational data results in the form of static line charts, bar charts, or dashboards. The data processing and display process primarily relies on database queries and fixed-format plotting to achieve a visual representation of HVAC operating status information.

[0004] Traditional HVAC data visualization relies on static graphic templates, failing to reflect real-time dynamic changes in heat and energy consumption. Updates depend on periodic refreshes, making it difficult to respond promptly to changes in environmental load. Under conditions of frequent load fluctuations, traditional line charts and dashboards struggle to accurately capture instantaneous fluctuations, lacking dynamic correlations of energy transfer and inter-node relationships, thus failing to effectively identify abnormal fluctuations and leading to delayed control responses. Furthermore, existing methods cannot display heat flow trends and cyclical changes over time, resulting in simplistic data presentation and a lack of intuitive trend visualization. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the existing technology and propose a visualization processing system for HVAC system operation data.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a HVAC system operation data visualization processing system includes:

[0007] The heat flux feature extraction module obtains heat flux density values ​​through a sensor network, uses a Kalman filter algorithm to filter out noise, analyzes node coordinates and orientation vector angles, generates a dataset of heat node graphic coordinates, and passes it to the orientation gradient difference calculation module.

[0008] The directional gradient difference calculation module receives the hot node graphic coordinate dataset, performs real-time difference calculation on the directional vector angle values ​​of adjacent nodes, identifies the directional change trend based on the difference sign, generates a rendering strategy for connecting adjacent nodes, and transmits it to the fluctuation amplitude analysis module.

[0009] The fluctuation amplitude analysis module receives the adjacent node connection rendering strategy, calculates the heat flux density fluctuation amplitude value according to the time series, analyzes the fluctuation trend direction and periodic change frequency using the sliding window averaging method, generates a node path reconstruction instruction, and transmits it to the connection attribute control module.

[0010] The connection attribute adjustment module calls the node path reconstruction instruction, performs a weighted operation on the directional gradient difference and color mapping coefficient to update the connection color parameters, performs a linear mapping calculation on the line width ratio coefficient and gradient difference to calculate the line width adjustment value, outputs the visualization map rendering result, and passes it to the dynamic trajectory optimization module.

[0011] As a further aspect of the present invention, the hot node graphic coordinate dataset includes a node spatial coordinate dataset, a direction vector angle matrix, and a set of heat flux density weighting coefficients; the adjacent node connection rendering strategy includes a set of direction change symbols, a set of node connection topology structures, and a set of rendering priority sequences; the node path reconstruction instructions include a set of fluctuation trend direction parameters, a set of periodic change frequency parameters, and a path coordinate adjustment matrix; and the visualization map rendering result includes a set of color mapping parameters, a set of line width ratio adjustment coefficients, and a set of layer display levels.

[0012] As a further aspect of the present invention, the heat flux feature extraction module includes:

[0013] The heat flux acquisition submodule acquires the heat flux density values ​​of the sampling nodes through the sensor network, sorts the sampling data according to the node time sequence number, detects the timestamp matching relationship and calculates the flux difference interval, compares the time delay according to the synchronization signal and reassembles the flux sequence to generate the node heat flux time series dataset.

[0014] The noise filtering submodule extracts the flux change rate of adjacent samples based on the node heat flux time series dataset, calculates the predicted and measured residual sequences, performs state prediction and observation update using the Kalman filter algorithm, adjusts abnormal samples and updates the index state, and obtains a smooth heat flux density sequence.

[0015] The node feature calculation submodule reads the node spatial coordinates and direction vectors based on the smoothed heat flux density sequence, calculates the distance vector and the angle difference, obtains the gradient difference based on the rate of change of direction and performs mapping, and obtains the thermal node graphic coordinate dataset.

[0016] As a further aspect of the present invention, the directional gradient difference calculation module includes:

[0017] The spatial data parsing submodule acquires the hot node graphic coordinate dataset, assigns topological numbers to the node coordinates, calculates the Euclidean distance difference based on the node index, detects the coordinate change rate according to the time step, calculates the displacement direction components and relative distance intervals between adjacent nodes, and generates a set of node displacement vectors.

[0018] The direction vector operation submodule extracts the coordinate difference of the direction components of adjacent nodes based on the node displacement vector set, calculates the unit direction vector angle, and performs linear interpolation when the angle difference is less than the angle reference threshold to generate a direction angle difference sequence.

[0019] The difference trend recognition submodule calculates the sign of the angle difference and detects the change range based on the direction angle difference sequence, counts the sign ratio, extracts the direction fluctuation frequency based on the positive and negative sign trends, and generates a rendering strategy for connecting adjacent nodes.

[0020] As a further aspect of the present invention, the method of detecting the rate of change of coordinates based on the time step refers to using adjacent time intervals in the time step sequence as a reference to detect the rate of change of node coordinates.

[0021] The angle reference threshold is the upper limit range of the direction angle variance calculated based on the direction distribution density of the node displacement vector set.

[0022] As a further aspect of the present invention, the fluctuation amplitude analysis module includes:

[0023] The heat flux calculation submodule calls the adjacent node connection rendering strategy, extracts the heat flux density change rate in the time series, compares the node transmission delay difference with the flux gradient difference, filters the time continuous difference sample set, calculates the heat flux density fluctuation characteristic value within the sliding window, and obtains the heat flux fluctuation amplitude result set.

[0024] The fluctuation trend analysis submodule, based on the heat flux fluctuation amplitude result set, divides the time window, extracts the fluctuation mean, calculates the sign change sequence of the mean difference between adjacent windows, determines the reversal node based on the number of sign switching, and counts the periodic change frequency to obtain the fluctuation trend spectrum vector.

[0025] The node path instruction generation submodule calls the fluctuation trend spectrum vector, filters the node sequence whose periodic change frequency is within the threshold range, calculates the node connection topology stability index and combines it with the frequency distribution characteristics to generate node path reconstruction instructions.

[0026] The threshold range is determined by analyzing the original fault data of the HVAC system, statistically analyzing the distribution of the fluctuation frequency of the fault occurrence nodes, and using the 10% quantile and 90% quantile as the lower and upper limits of the threshold range.

[0027] As a further aspect of the present invention, the connection attribute control module includes:

[0028] The directional gradient difference calculation submodule calls the node path reconstruction instruction to obtain the direction vector data of the connection between nodes, extracts the node coordinate sequence, calculates the angle change based on the coordinate difference, performs a difference and offset ratio judgment on the node direction change, arranges the direction offset ratio, and obtains the directional gradient difference result.

[0029] The color parameter weighted update submodule extracts node color channel parameters based on the directional gradient difference result, calculates the corresponding change of gradient difference and color coefficient, summarizes the differences in node color parameters, calculates update parameters and maps them to the connection color channel to form the connection color parameter result.

[0030] The linewidth mapping adjustment submodule extracts the gradient difference of the connection lines based on the line color parameter results, calculates the linear mapping amount between the scaling factor and the gradient difference, analyzes the gradient of linewidth change and superimposes the original linewidth sequence, and outputs the visualization map rendering result.

[0031] As a further aspect of the present invention, the color mapping update parameters are calculated using the following formula:

[0032] ;

[0033] in, This represents the color mapping update parameter. This represents the gradient value in the direction of the i-th node. Represents the reference value of the directional gradient. represents the color mapping coefficient of the i-th node, and m represents the total number of nodes.

[0034] As a further embodiment of the present invention, the dynamic trajectory optimization module receives the visualization map rendering result, calculates the coordinate offset based on the fluctuation amplitude value and displacement coefficient, updates the coordinate position of the node in the map according to the fluctuation trend direction, performs real-time binding calculation of trajectory tailing parameters and current coordinates, and generates a dynamic trajectory display effect.

[0035] The dynamic trajectory display effect includes a coordinate offset sequence set, a node motion trajectory dataset, and a trailing display control parameter set.

[0036] As a further aspect of the present invention, the dynamic trajectory optimization module includes:

[0037] The rendering data parsing submodule obtains the rendering result of the visualization map, collects the node fluctuation amplitude value and displacement coefficient, detects the data integrity and removes abnormal items, performs proportional conversion based on the fluctuation amplitude value and displacement coefficient, and generates the node coordinate offset.

[0038] The displacement calculation submodule establishes an initial coordinate set of nodes based on the node coordinate offset, superimposes the initial coordinate set of nodes with the offset base amount, extracts the fluctuation trend direction parameter and performs sign matching to adjust the coordinate offset direction, updates the real-time coordinate position of the nodes, and obtains the dynamic displacement of the nodes.

[0039] The offset base quantity is a unit direction vector, with the initial direction set along... Positive direction of the axis;

[0040] The trajectory generation submodule calls the dynamic displacement of the node to establish a node coordinate update sequence, arranges it into a displacement time chain in chronological order, binds the trajectory tail parameter with the current coordinate of the node, and generates a dynamic trajectory display effect.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0042] In this invention, by real-time filtering of heat flux density and calculation of direction vectors, dynamic correlation mapping of heat transfer paths between multiple nodes can be performed, ensuring the continuity of heat flow changes and avoiding the delays and distortions in data updates inherent in traditional static graphics. Time series smoothing is used to accurately capture heat flow trends and periodic changes; fluctuation amplitude and direction can be intuitively presented through color and line width variations, demonstrating changes in heat flow intensity and direction. This approach enhances the temporal awareness of thermal environment data, improves trend recognition and dynamic response capabilities, and clearly displays changes in the energy transfer process, effectively addressing the shortcomings of traditional methods in real-time performance and spatial relationship representation. Attached Figure Description

[0043] Figure 1 This is a system flowchart of the present invention;

[0044] Figure 2 This is a flowchart illustrating the acquisition process of the heat flux feature extraction module in this invention.

[0045] Figure 3 This is a flowchart illustrating the acquisition process of the directional gradient difference calculation module in this invention.

[0046] Figure 4 This is a flowchart illustrating the acquisition process of the fluctuation amplitude analysis module in this invention.

[0047] Figure 5 This is a flowchart illustrating the acquisition process of the connection attribute control module in this invention.

[0048] Figure 6 This is a flowchart illustrating the acquisition process of the dynamic trajectory optimization module in this invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the 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 illustrative and not intended to limit the invention.

[0050] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0051] Please see Figure 1 A data visualization and processing system for HVAC system operation includes:

[0052] The heat flux feature extraction module obtains heat flux density values ​​through a sensor network, uses a Kalman filter algorithm to filter out noise, analyzes node coordinates and orientation vector angles, generates a dataset of heat node graphic coordinates, and passes it to the orientation gradient difference calculation module.

[0053] The directional gradient difference calculation module receives the hot node graphic coordinate dataset, performs real-time difference calculation on the directional vector angle values ​​of adjacent nodes, identifies the directional change trend based on the difference sign, generates a rendering strategy for connecting adjacent nodes, and transmits it to the fluctuation amplitude analysis module.

[0054] The fluctuation amplitude analysis module receives the rendering strategy for the connection between adjacent nodes, calculates the fluctuation amplitude value of heat flux density according to the time series, analyzes the fluctuation trend direction and periodic change frequency using the sliding window averaging method, generates node path reconstruction instructions, and transmits them to the connection attribute control module.

[0055] The connection attribute adjustment module calls the node path reconstruction instruction, performs a weighted calculation on the directional gradient difference and color mapping coefficient to update the connection color parameters, performs a linear mapping calculation on the line width ratio coefficient and gradient difference to calculate the line width adjustment value, outputs the visualization map rendering result, and passes it to the dynamic trajectory optimization module.

[0056] The dynamic trajectory optimization module receives the rendering results of the visualization map, calculates the coordinate offset based on the fluctuation amplitude value and displacement coefficient, updates the coordinate position of the node in the map according to the fluctuation trend direction, performs real-time binding calculation of trajectory tailing parameters and current coordinates, and generates dynamic trajectory display effect.

[0057] The hot node graphical coordinate dataset includes a node spatial coordinate dataset, a direction vector angle matrix, and a set of heat flux density weighting coefficients. The adjacent node connection rendering strategy includes a set of direction change symbols, a set of node connection topology structures, and a set of rendering priority sequences. The node path reconstruction instructions include a set of fluctuation trend direction parameters, a set of periodic change frequency parameters, and a path coordinate adjustment matrix. The visualization map rendering results include a set of color mapping parameters, a set of line width ratio adjustment coefficients, and a set of layer display levels. The dynamic trajectory display effect includes a set of coordinate offset sequences, a node motion trajectory dataset, and a set of trailing display control parameters.

[0058] Please see Figure 2 The heat flux feature extraction module includes:

[0059] The heat flux acquisition submodule acquires the heat flux density values ​​of the sampling nodes through the sensor network, sorts the sampling data according to the node time sequence number, detects the timestamp matching relationship and calculates the flux difference interval, compares the time delay according to the synchronization signal and reassembles the flux sequence to generate the node heat flux time series dataset.

[0060] The raw heat flux density data acquired from the sensor network comprises multiple sampling nodes, specifically a series of discrete data points including node numbers, timestamps, and heat flux density values. For example, this data can be obtained from three heat flux sensors installed on the outer wall of HVAC system pipes (node ​​numbers are...). In continuous The raw data collected within seconds was initially unordered, as shown in Table 1. First, based on the time sequence number (timestamp) of each data point, the sampled data was sorted in ascending order. exist The data points collected at all times are placed exist Before each data point, a strictly time-evolving data sequence is formed. Then, the timestamp matching relationships are checked against the sorted sequence, and a timestamp matching tolerance is set. This tolerance is determined based on the precision of the system's synchronization clock, for example, set to [value missing]. The millisecond tolerance value is set with reference to the maximum data packet transmission jitter time defined in the sensor network communication protocol. The specific setting process is as follows: statistical continuous... Within each synchronization signal cycle, the difference between the actual arrival time and the theoretical arrival time of node data quantile, then multiplied by one The safety factor, if the difference within a certain period quantiles are If the tolerance is milliseconds, then it is set to... milliseconds, rounded to the nearest integer Milliseconds, iterate through the sequence, check the difference between the timestamps of any two adjacent data points, if the difference is less than... A millisecond is considered a matching pair, and then the difference in heat flux density values ​​is calculated for each matching pair. For example, in the sequence... exist The value at time , exist The value at time The difference between the timestamps is milliseconds, less than Milliseconds form matching pairs, and their flux difference is The calculated differences are collected, and the maximum and minimum values ​​are determined to obtain the flux difference range for the entire sequence. Then, an independent system-level synchronization signal sequence is introduced, which emits a precise timestamp every second, such as... The time delay is calculated by comparing the timestamp of each sampled data point with the timestamp of the nearest synchronization signal. exist The most recent synchronization signal of the sampled data is The time delay is milliseconds, if exist Sampling, the most recent synchronization signal is The delay is In milliseconds, a time delay threshold is set, which is based on network device performance metrics and set to [value missing]. Milliseconds, the time delay of a data point, such as If the value exceeds this threshold in milliseconds, it is considered delayed data and determined based on its timestamp. Find the insertion position again in the sorted sequence and move it to the timestamp. After collecting the data points, the data is finally reorganized to generate a node heat flux time series dataset.

[0061] Table 1: Raw Heat Flux Monitoring Data

[0062] ;

[0063] As shown in Table 1, the table lists the unprocessed raw data collected from three different sensor nodes, including node identifiers, timestamps accurate to milliseconds, and corresponding measured values ​​of heat flux density.

[0064] The noise filtering submodule extracts the flux change rate of adjacent samples based on the node heat flux time series dataset, calculates the predicted and measured residual sequences, performs state prediction and observation update using the Kalman filter algorithm, adjusts abnormal samples and updates the index state, and obtains a smooth heat flux density sequence.

[0065] Based on the generated node heat flux time series dataset, such as a dataset composed of... The sequence consists of n data points. First, the flux change rate between adjacent sampling points is extracted from this dataset. Specifically, for the nth data point in the sequence... Data points (Its timestamp is) ), compared to the previous data point (timestamp is) ) calculate the rate of change For example, if a data point in the dataset is (exist (Time), the next data point is (exist At time ( ), the rate of change of flux between the two points is . Then, based on this rate of change, the predicted value for the current moment is calculated. The smoothed value from the previous time step With process noise The decision, and its calculation process, are as follows:

[0066] ;

[0067] Assumption The smoothing value at time step ,but The predicted value at time is Next, the predicted value is compared with the actual measured value at the current moment. The residual between, i.e. Substitute the values, This forms a residual sequence. Then, state prediction and observation update are performed, first calculating a gain coefficient. The calculation of this coefficient involves the prediction error covariance. Covariance of measurement noise ,in Determined based on the accuracy specified by the sensor at the factory, for example, ,and The error covariance from the previous time step Add process noise covariance We get, assuming , ,but The specific calculation of the gain coefficient is as follows: Using this gain coefficient and residual To update the predicted values ​​and obtain the smoothed values ​​at the current time. Substitute the values, Simultaneously update the error covariance:

[0068] ;

[0069] Next, outlier samples are adjusted based on the residual sequence, and an outlier discrimination interval is defined, the boundary of which is... ,in Therefore, the interval is If the current residual If the sample is located within this interval, then the sample If the residual of another sample is considered normal, it is considered normal. If the value is not found, it is considered an anomalous sample. For anomalous samples, the corresponding smoothing value is... Replace the original measurement value The position of the data point in the sequence is determined, and the index status of that data point is updated, for example, by adding a "corrected" flag. This process is repeated for all data points in the time series dataset to obtain a smoothed heat flux density sequence.

[0070] The node feature calculation submodule reads the node spatial coordinates and direction vectors based on the smoothed heat flux density sequence, calculates the distance vector and the angle difference, calculates the gradient difference based on the rate of change of direction and performs mapping, and obtains the thermal node graphic coordinate dataset.

[0071] Based on the output smoothed heat flux density sequence, which is a series of timestamps and corresponding smoothed heat flux density values, such as nodes... At any moment The smoothing value is ,node At any moment The smoothing value is First, it reads the preset three-dimensional spatial coordinates and surface normal direction vectors of multiple sampling nodes from the system configuration library, for example, The coordinates are Meters, the surface normal direction vector is , The coordinates are meters, its direction vector is Next, the angle difference between the spatial distance vector and the direction vector between any two nodes is calculated. The distance vector is obtained by subtracting the coordinates. Meters, the distance is:

[0072] rice;

[0073] Two direction vectors and The angle difference between This is obtained through vector dot product operation, specifically by calculating... The magnitudes of both vectors are Therefore, the included angle Then, the gradient difference is calculated based on the rate of change of direction between nodes. This gradient difference is defined as the smooth difference in heat flux density between two nodes divided by the straight-line distance between them, i.e.:

[0074] ;

[0075] This gradient difference will be compared with a preset gradient benchmark value, which is set according to the allowable upper limit of heat loss in pipes in the HVAC system design code. For example, for an insulation layer of this pipe diameter, the maximum allowable change in heat flux density per unit distance is... The calculated gradient difference If the value is less than this benchmark, it indicates that the heat distribution is within the normal range. Finally, the spatial coordinates of the thermodynamic nodes are mapped from the three-dimensional physical space coordinates. Convert to 2D graphic display coordinates The mapping process uses a fixed oblique axonometric projection transformation, for example, the transformation rule is as follows: and , will node coordinates Substitute them to obtain its graphical coordinates. , This transformation is applied to the nodes, ultimately generating a dataset containing the two-dimensional graphical coordinates of the nodes and their corresponding smoothed heat flux density values, i.e., the thermal node graphical coordinate dataset.

[0076] Please see Figure 3 The directional gradient difference calculation module includes:

[0077] The spatial data parsing submodule acquires the hot node graphic coordinate dataset, assigns topological numbers to the node coordinates, calculates the Euclidean distance difference based on the node index, detects the rate of coordinate change based on the time step, calculates the displacement direction components and relative distance intervals between adjacent nodes, and generates a set of node displacement vectors.

[0078] Obtain the generated hot node graphical coordinate dataset, which contains a series of two-dimensional graphical coordinates of nodes at different timestamps, as shown in Table 2. First, topologically number the node coordinates in the dataset, specifically by assigning a fixed integer index to each unique node. For example, for nodes... Number ,node Number Next, based on the node index, at the same timestamp Internally calculate any two adjacent nodes, for example Node No. and Node number, its coordinates and The difference in Euclidean distance between them is calculated by substituting the numerical values: Distance:

[0079] ;

[0080] Then, the rate of change of the coordinates of a single node is detected based on a fixed time step. Let the time step be... for seconds, i.e., the data acquisition frequency is ,for Node number, at The coordinates of the time are ,exist The coordinates of the time are Then its coordinate change is , Its coordinate change rate is m / s, then calculate the displacement direction components and relative distance intervals between adjacent nodes, for Number and Node number, at time step The displacement direction component inside is and At the same time, the adjacent node pairs are calculated. The distance at any given time, and the minimum and maximum distance values ​​are used as the relative distance interval. For example, if there still exists... Node number 1, and calculate The relative distance interval is Finally, the coordinate change rate vectors calculated for each node in each time step are collected to generate a set of node displacement vectors.

[0081] Table 2: Time Series Data of Hot Node Graphical Coordinates

[0082] ;

[0083] As shown in Table 2, this table lists the two hot nodes at two consecutive time steps. and The two-dimensional graphic coordinates below are the basis for displacement and direction analysis.

[0084] The direction vector operation submodule extracts the coordinate difference of the direction components of adjacent nodes based on the node displacement vector set, calculates the unit direction vector angle, and performs linear interpolation when the angle difference is less than the angle reference threshold to generate a direction angle difference sequence.

[0085] Based on the generated set of nodal displacement vectors, which contains the displacement components of nodes per unit time, for example, node... The displacement vector is Adjacent nodes The displacement vector is First, extract the component coordinate difference of the displacement vectors of these two adjacent nodes, that is... Next, calculate the unit direction vector angle for each of the two displacement vectors. The angle is calculated as the ratio of the arctangent function acting on the vertical component to the horizontal component. For nodes... Its angle is For nodes Its angle is Then calculate the difference between these two angles. This angle difference is compared with a preset angle reference threshold. This threshold is set with reference to the maximum streamline deflection angle corresponding to the critical Reynolds number in the transition zone between laminar and turbulent flow in fluid mechanics. By consulting relevant HVAC duct design manuals, this angle is set at a specific value for a given flow velocity and pipe diameter. The specific setup process involves collecting a large amount of displacement data of adjacent nodes under stable operating conditions, calculating the probability distribution of their angle differences, and then taking their values. The upper bound of the confidence interval is used as the benchmark value. Let's assume the benchmark value calculated using this method is... Due to the currently calculated angle difference Less than the reference threshold for this angle If the displacement direction of these two nodes changes smoothly, then a linear interpolation is performed on the angle between them. The purpose of the interpolation is to generate an intermediate transition direction during visualization rendering. The interpolation operation involves creating a virtual point between the two nodes with a direction angle of... If the angle difference is greater than If the angle difference is not calculated, the interpolation operation is not performed. Instead, the angle difference is recorded directly, and the angle difference values ​​calculated for adjacent node pairs (regardless of whether interpolation is performed) are combined in order to generate a direction angle difference sequence.

[0086] The difference trend recognition submodule calculates the sign of the angle difference and detects the change range based on the direction angle difference sequence, counts the sign ratio, extracts the direction fluctuation frequency based on the positive and negative sign trends, and generates a rendering strategy for connecting adjacent nodes.

[0087] Based on the generated sequence of direction angle differences, for example, a sequence containing five consecutive calculated values. First, the sign of each angle difference in the sequence is calculated to obtain a sign sequence. It detects the intervals where the sign changes, i.e., the positions from positive to negative or from negative to positive, in the sequence where the sign is at the th . The and the first Between elements, the first The and the first Between elements, the first The and the first The elements changed, and then the ratio of positive to negative signs in the entire sequence was counted. Positive signs appeared. Next, the negative sign appeared. times, totaling Therefore, the proportion of positive signs is [number], The proportion of negative signs is Next, based on the alternating trend of the positive and negative sign sequences, the directional fluctuation frequency is extracted. This frequency is defined as the number of sign changes divided by the total length of the sequence minus one. In the middle, the sign changed. The sequence length is [number]. Therefore, the fluctuation frequency is The fluctuation frequency is compared with a preset fluctuation range to determine the degree of fluctuation. The fluctuation range is determined based on the analysis of the system's original data, and fluctuation frequencies below a certain threshold are considered normal. The definition is low volatility, between and The range between is defined as medium volatility, and above The definition is high volatility, and the currently calculated frequency is... Belonging to the high volatility range, finally, based on the sign ratio and volatility frequency results, a rendering strategy for connecting adjacent nodes is generated. Specifically, if the volatility frequency is rated as high, the line segment connecting the two nodes is rendered as a red dashed line; if the volatility frequency is medium, it is rendered as a yellow solid line; and if the volatility frequency is low, it is rendered as a green solid line. Furthermore, if the positive sign ratio exceeds... Then, add an arrow pointing from the previous node to the next node on the line segment. Based on the current example result, the fluctuation frequency is... (High volatility), with a positive sign ratio of (Exceed Therefore, the final rendering strategy is to connect the two nodes with a red dashed line with an arrow.

[0088] Please see Figure 4 The fluctuation amplitude analysis module includes:

[0089] The heat flux calculation submodule calls the adjacent node connection rendering strategy, extracts the heat flux density change rate in the time series, compares the node transmission delay difference with the flux gradient difference, and filters out the time-continuous difference sample set using the formula:

[0090] ;

[0091] Calculate the characteristic value of heat flux density fluctuation within the sliding window to obtain the result set of heat flux fluctuation amplitude;

[0092] in, This represents the characteristic value of the heat flux density fluctuation at the i-th node at time t. This represents the heat flux density of the i-th node at the k-th sampling point at time t. This represents the heat flux density at the k-th sampling point of the i-th node at time t-1. Represents the time interval between adjacent sampling points. This represents the number of sampling points within the sliding window;

[0093] Invoke the rendering strategy for the generated adjacent node connections, for example, when a node... and When the connection between the two nodes is rendered as a red dashed line, the calculation of the heat flux data for these two nodes is immediately triggered. First, the time-series data associated with these two nodes is extracted from the smoothed heat flux density sequence, specifically at the current time. and the previous moment The sampled heat flux density values ​​for each node within two time slices are shown in Table 3. Next, the data streams of these two nodes are compared, and the transmission delay difference between their data packets arriving at the server is calculated. For example... exist The data packet delay at time is millisecond, for milliseconds, latency difference is Milliseconds, while simultaneously calculating the obtained flux gradient difference, its value is Then, a filtering condition is set, which requires that the transmission delay difference must be less than 1 / 3. Milliseconds, and the flux gradient difference must be greater than [a certain value]. At that time, the delay threshold The millisecond setting references the typical accuracy of NTP time synchronization within a local area network, while the flux gradient difference threshold... This is based on the heat dissipation rate under normal operating conditions in the HVAC system. The quantiles are determined by the current calculation. Millisecond latency difference is less than milliseconds, and The gradient difference is greater than This sample set was identified as a time-continuous difference sample set and selected for subsequent calculations. The characteristic value of heat flux density fluctuation within the sliding window was then calculated using the following formula:

[0094] ;

[0095] This formula is used to calculate within a time window. Inside, node The root mean square value of the rate of change of heat flux density, where Representing the Each node A quantitative indicator of the magnitude of fluctuations at any given moment. yes The first time within the sliding window The heat flux density value at each sampling point It was the previous moment The value corresponding to the sampling point, It is the time interval between two adjacent sampling points, and This represents the total number of sampling points contained within the sliding window. By squaring the absolute value of the difference, the impact of drastically changing sampling points on the result is amplified. The square root operation ensures that the dimensions of the result are consistent with the rate of change. The advantage of this formula is that by calculating the quadratic average of the rate of change, it can more sensitively capture instantaneous and drastic fluctuations in heat flux, rather than averaging them out by gradual changes. For nodes... (Right now Set the size of the sliding window. Obtain the data from Table 3 and perform the following calculations:

[0096] ;

[0097] The calculation result Indicate node exist The characteristic value of heat flux density fluctuation at time t is This value will be added to a set, and this calculation will be performed on all nodes that meet the conditions, ultimately yielding a set of heat flux fluctuation results.

[0098] Table 3: Nodal Heat Flux Density Sampling Data

[0099] ;

[0100] As shown in Table 3, the table lists the nodes used for computation. exist The specific sampling data of the time-fluctuation characteristic value includes three sets of corresponding heat flux density values ​​and sampling time intervals under two adjacent time stamps.

[0101] The fluctuation trend analysis submodule, based on the heat flux fluctuation amplitude result set, divides the time window, extracts the fluctuation mean, calculates the sign change sequence of the mean difference between adjacent windows, determines the reversal node based on the number of sign switching, and statistically analyzes the periodic change frequency to obtain the fluctuation trend spectrum vector.

[0102] The calculated heat flux fluctuation amplitude result set consists of a series of fluctuation characteristic values ​​that vary over time, for example, for nodes. The sequence of fluctuation eigenvalues ​​obtained over six consecutive time steps is as follows: First, divide this time series into time windows of equal length, and set the window size to [value missing]. With a time step of 1, the sequence is divided into two windows, and the window... and window Next, the arithmetic mean of the fluctuation feature values ​​within each window is extracted. Its mean fluctuation is ,for Its mean fluctuation is Then, calculate the difference between the means of adjacent windows and record its sign; the difference is... Its sign is positive, and this process is applied to a longer time series, such as continuous time series. The mean sequence of the windows is as follows:

[0103] ;

[0104] The mean difference between its adjacent windows is as follows: This yields a sequence of mean difference sign changes. The number of inverted nodes is determined based on the number of sign switching events in this symbol sequence. In this sequence, a sign switching from positive to negative occurs once, from negative to positive occurs once, and from positive to negative occurs again once, for a total of [number missing]. This switch The point in time where the switching occurs is the reversal point of the fluctuation trend. Finally, the frequency of the periodic changes of the reversal over the entire observation period is statistically analyzed, assuming that this... The total duration covered by the time windows is If the periodic change frequency is seconds, then the frequency of the periodic change is The second switch divided by Seconds, that is The frequency value calculated for each node is used as an element representing the core characteristics of its fluctuation trend, and combined into a multi-dimensional vector to obtain the fluctuation trend spectrum vector.

[0105] The node path instruction generation submodule calls the fluctuation trend spectrum vector, filters the node sequence whose periodic change frequency is within the threshold range, calculates the node connection topology stability index and combines it with the frequency distribution characteristics to generate node path reconstruction instructions.

[0106] The generated fluctuation trend spectrum vector is invoked, which contains the periodic change frequency of the nodes, such as the node... The frequency is ,node The frequency is First, node sequences with periodic variation frequencies within a preset threshold range are selected. This threshold range is designed to identify nodes whose fluctuations are neither too frequent (caused by random noise) nor too infrequent (belonging to normal state changes). Specifically, the process involves analyzing the original fault data of the HVAC system and statistically analyzing the data before the fault occurred. The distribution of node fluctuation frequency within an hour, using the 10th percentile and 90th percentile as the lower and upper limits of the threshold interval, is assumed to be set accordingly. ,node frequency Located within this range, it is therefore selected, and the node... frequency Nodes not within the specified interval are excluded, resulting in a sequence of nodes to be processed. Next, for each node in the sequence, the topological stability index of its connections with neighboring nodes is calculated. This index is composed of two weighted components: one is the relative distance between nodes within a continuous interval... The variance within each time step is one part, and the mean of its direction angle difference sequence is another part, with weights set as follows: and The weight values ​​were determined through sensitivity analysis on the original dataset. By testing the impact of different weight combinations on the accuracy and recall of the instability risk prediction model, it was found that the model's overall performance was optimal when the distance stability weight was 0.6 and the directional stability weight was 0.4. This result quantitatively indicates that distance stability is slightly more important than directional stability when evaluating topological relationships, for example, when computing nodes. with his neighbors The stability of the distance variance is... The average difference in direction angles is The stability index is Then, this stability index is combined with the frequency distribution characteristics of the nodes for judgment. A judgment rule is set: if the frequency of a node is within a threshold range and the average value of the topology stability index of its connections is lower than a stability benchmark value, then an instruction is triggered to generate. This stability benchmark value is set as the node stability index under normal operating conditions. quantile, assuming the baseline value is Due to the node frequency Within the interval, and its average stability Below If the path of that node is deemed unstable, a path reconstruction instruction is generated for that node. This instruction adjusts the node connections in the data visualization; for example, the instruction could be "disconnect node". and The connection will The connection target is switched to the neighboring node with the highest topological stability. ".

[0107] Please see Figure 5 The connection attribute control module includes:

[0108] The directional gradient difference calculation submodule calls the node path reconstruction instruction to obtain the direction vector data of the connection between nodes, extracts the node coordinate sequence, calculates the angle change based on the coordinate difference, performs a difference and offset ratio judgment on the node direction change, arranges the direction offset ratio, and obtains the directional gradient difference result.

[0109] Invoke the generated node path reconstruction command, for example, the command content is "disconnect node". and The connection will The connection target is switched to the neighboring node with the highest topological stability. First, retrieve the newly established connections between nodes after executing the instruction from the system. arrive The direction vector data is then extracted from these two nodes at two consecutive time steps. and The sequence of graphic coordinates, for example, in time, The coordinates are , The coordinates are ,exist time, The coordinates become , The coordinates become The angle change of the connecting line direction is calculated based on the coordinate difference. At time t, the connecting vector is Its direction angle is ,exist At time t, the connecting vector is Its direction angle is Then in The change in angle during this period is Then, the difference and offset ratio are judged for the node direction changes calculated in multiple consecutive time steps, assuming that in the previous time step The change in angle inside is The difference in the current change is Set an offset ratio judgment criterion, which is the absolute value of the current angle change. With the past Compare the average of the absolute values ​​of the angle change within each time step. If the average is... The offset ratio is This ratio is compared with an offset threshold. For comparison, this threshold is derived from the maximum value of the statistical distribution of the offset ratio during 95% of the system's stable operation time. Less than If the change in direction is normal, then the offset ratios calculated by connecting the nodes are sorted in descending order to form an ordered list, and the directional gradient difference result is obtained.

[0110] The color parameter weighted update submodule extracts the node color channel parameters based on the directional gradient difference results, calculates the corresponding changes in the gradient difference and color coefficients, and summarizes the differences in node color parameters using the following formula:

[0111] ;

[0112] Calculate the color mapping update parameters and map them to the color channels of the connection to form the result of the connection color parameters;

[0113] in, This represents the color mapping update parameter. This represents the gradient value in the direction of the i-th node. Represents the reference value of the directional gradient. represents the color mapping coefficient of the i-th node, and m represents the total number of nodes;

[0114] Based on the calculated directional gradient difference, which is an ordered list containing the offset ratios of node connection directions, the color channel parameters currently associated with the nodes in the list are first extracted from the visualization configuration library. For example, the node... The color mapping coefficient is ,node for ,node for ,node for As shown in Table 4, the changes in directional gradient difference and preset color coefficients are then calculated, and the differences in color parameters of the nodes are summarized using the formula:

[0115] ;

[0116] Calculate the updated parameters of the color mapping in this formula. It is a comprehensive indicator used to adjust colors. It is the first The directional gradient difference (i.e., directional offset ratio) is calculated from each node. It is a directional gradient baseline value, representing the average gradient level of the system in a steady state. It is the first The color mapping coefficients of each node determine the weight of the gradient bias of that node in the final color update. This is the total number of nodes involved in the calculation. This formula adds the weighted sum of gradient biases to the Euclidean norm of the coefficient vector, then normalizes it using the number of nodes and the sum of absolute biases. The advantage of this formula is that by introducing a square root term for the sum of squares of the color mapping coefficients, it ensures that even when gradient values ​​are close to the baseline, the updated parameters still retain a fundamental value determined by the coefficients themselves, avoiding complete stagnation in the update process. (Setting the total number of nodes...) directional gradient reference value The average value was obtained by statistically analyzing the original stable data and set as [value]. The gradient differences of the nodes are respectively Substitute the values ​​into the calculation:

[0117] Calculation of the molecule, Part 1:

[0118] ;

[0119] Calculation of the molecule, Part 2:

[0120] ;

[0121] Calculate the denominator:

[0122] ;

[0123] final:

[0124] ;

[0125] This result indicates that the current system's overall color update parameter is: This parameter will act as a multiplier or addend, uniformly affecting the color channels of the reconstructed connection. For example, if the base color of the original connection is determined by the hue value in the HSV model... (Green) indicates that the updated color tone is The color shifts towards yellowish-green, resulting in the color parameter result of the connecting line.

[0126] Table 4: Node Color Mapping Coefficients

[0127] ;

[0128] As shown in Table 4, the table provides the preset color mapping coefficient values ​​for the four nodes involved in the color update calculation. The coefficients determine the weight of the impact of node state changes on visual presentation.

[0129] The line width mapping adjustment submodule extracts the gradient difference of the connection based on the line color parameter results, calculates the linear mapping amount between the scale coefficient and the gradient difference, analyzes the gradient of line width change and superimposes the original line width sequence, and outputs the visualization map rendering result.

[0130] Based on the generated connection color parameter results, the colors of the associated connections have been updated according to the parameters. A unified adjustment was made. First, the directional gradient difference corresponding to the adjusted connection was extracted from the system. For example, for the connection... Its gradient difference is Next, calculate the linear mapping between a scaling factor and the gradient difference, setting the base line width to... The pixel and line width variation coefficient is The setting of this coefficient is based on UI design principles, ensuring that line width changes are visually clear and discernible without being too abrupt. The specific setting process involves rendering lines of different widths on the screen, from at least... The testers observed and selected those who were... At a viewing distance of centimeters, the smallest linewidth increment that can be generally recognized as a "significant change" is obtained by dividing this increment by the average of the corresponding gradient differences. The process involves calculating the current linewidth increment (pixel). Then, the current linewidth change gradient is analyzed and superimposed on the original linewidth sequence. Here, linewidth change gradient analysis refers to combining the currently calculated linewidth increment... With a preset maximum allowed increment For comparison, the maximum increment value is set to the base line width. ,Right now Pixels, due to the calculated increment Less than The increment is considered smooth and is adopted; if the calculated increment is greater than... Then reduce its value to This mapping value is superimposed on the original line width of the connection to obtain the new line width. For each pixel, this process is repeated for the lines that need updating, and the updated color and line width parameters are sent to the rendering engine to output the visual graph rendering results.

[0131] Please see Figure 6 The dynamic trajectory optimization module includes:

[0132] The rendering data parsing submodule obtains the visualization map rendering results, collects the node fluctuation amplitude value and displacement coefficient, checks the data integrity and removes anomalies, performs proportional conversion based on the fluctuation amplitude value and displacement coefficient, and generates node coordinate offset.

[0133] The output visualization rendering result is obtained, which includes the updated node connection color and line width information. First, the latest node fluctuation amplitude value and displacement coefficient corresponding to the node are collected from the system. The fluctuation amplitude value comes from the heat flux fluctuation amplitude result set, for example, node... The fluctuation range value The displacement coefficient is a parameter preset during system initialization based on the physical importance or monitoring sensitivity of each node, as shown in Table 5. Next, the collected data undergoes an integrity check. This check involves verifying each node's data record one by one to confirm whether it simultaneously contains the four fields: timestamp, node number, fluctuation amplitude value, and displacement coefficient. If any field is missing, the record is considered incomplete. For example, if a node's data is collected... If the data is complete but lacks fluctuation range values, the record will be removed and will not participate in subsequent calculations. After confirming the data is complete, outliers will be removed. The criteria for judging outliers is that if the fluctuation range value of a certain node is missing... Exceeded the past of this node The average amplitude value within each time step If the value is multiple times higher, it is considered an anomaly caused by transient interference from the sensor, for example, at a node. past The average amplitude of each time step is ,That times Current amplitude value Since the data did not exceed this upper limit, it was determined to be normal data. Subsequently, a proportional conversion was performed based on the normal fluctuation amplitude value and displacement coefficient. This conversion aims to transform the physical quantity (fluctuation amplitude) into a visual offset in the graphical interface. The conversion process involves multiplying the node's fluctuation amplitude value by its corresponding displacement coefficient, and then multiplying it by a global scaling factor, which is set to [value missing]. Its purpose is to control the maximum offset distance of nodes on the screen, avoiding screen chaos caused by excessive fluctuations in individual nodes. For example, its fluctuation range is The displacement coefficient is Then the magnitude of its coordinate offset is Each pixel unit is used to collect the calculated offset values ​​of the nodes and generate the node coordinate offset.

[0134] Table 5: Table of Nodal Displacement Coefficients

[0135] ;

[0136] As shown in Table 5, this table lists the displacement coefficient values ​​used by the main monitoring nodes when performing displacement conversion. This coefficient determines the degree of visual offset of different nodes under the same fluctuation amplitude.

[0137] The displacement calculation submodule establishes an initial coordinate set of nodes based on node coordinate offset, superimposes the initial coordinate set of nodes with the offset base amount, extracts the fluctuation trend direction parameter and performs sign matching to adjust the coordinate offset direction, updates the real-time coordinate position of nodes, and obtains the dynamic displacement of nodes.

[0138] Based on the generated node coordinate offset, this offset is a scalar value representing the distance the node needs to be offset, for example, node... The offset size is First, an initial coordinate set for the nodes is established based on the system configuration. This coordinate set represents the static reference position of the nodes when no offset occurs. For example, node... The initial coordinates are Next, the initial coordinate set of the nodes is superimposed with a base offset, where the base offset is a unit direction vector, and its initial direction is set along... The positive direction of the axis, i.e. , offset size Multiplying by this unit vector yields an initial offset vector. This is then superimposed on the initial coordinates to obtain a temporary coordinate system. Subsequently, the fluctuation trend direction parameter of the node is extracted from the results of the fluctuation trend analysis submodule. This parameter is a symbol representing the increase or decrease trend of fluctuation. If the mean difference of the window is positive ("+"), sign matching is needed to adjust the coordinate offset direction. The matching rule is: if the sign is "+", then the initial offset vector... Rotate counterclockwise around the origin If the symbol is "-", then rotate clockwise. The rotation operation is achieved by multiplying by a rotation matrix for a node. The positive sign, its offset vector is adjusted to Finally, this oriented offset vector is applied to the node's initial coordinates to update the node's real-time coordinate position. The final real-time coordinates are The real-time coordinates of the nodes after the updates are summarized to obtain the dynamic displacement of the nodes.

[0139] The trajectory generation submodule calls the dynamic displacement of nodes to establish a node coordinate update sequence, arranges it into a displacement time chain in chronological order, binds the trajectory tailing parameter with the current coordinates of the nodes, and generates a dynamic trajectory display effect.

[0140] The calculated dynamic displacement of the nodes is used, which represents the final displayed coordinates of multiple nodes at the current timestamp. First, a coordinate update sequence is established for each node. This sequence is a first-in-first-out queue data structure used to store the original coordinates of the nodes over the most recent few time steps. For example, for node... Its current coordinates are Assuming its coordinates in the first two time steps are respectively and Then its coordinate update sequence is:

[0141] ;

[0142] This sequence is automatically arranged chronologically to form a displacement time chain. Then, preset trajectory tail parameters are bound to the current coordinates of the nodes. The trajectory tail parameters are a set of configuration values ​​used to control visual effects, including the tail length, i.e., the queue length. The maximum length is set to Each coordinate point has a trail start color set to the node's current color value, a trail end color set to completely transparent, a trail start width set to 80% of the width of the current node's connecting line, and a trail end width set to... The pixel settings referenced the trajectory design specifications of mainstream industry visualization tools and standard libraries (such as D3.js, Three.js, ECharts, Tableau, and Cesium dynamic dot effects), while also incorporating empirical research results on user visual perception. Length standard: In D3.js and Cesium path animations, the generally recommended tail length is 5–15 sampling points. Too short a length will cause discontinuous trajectories, while too long a length will lead to visual overlap and performance degradation. Therefore, a middle value of 10 was chosen as a robust solution. Color gradient standard: ECharts and Three.js recommend using a linear gradient from "solid color → transparent" to highlight the direction of motion and avoid trajectory afterimages interfering with static nodes; interpolation in HSL space can also avoid abrupt color changes. Width gradient standard: Tableau and D3 suggest that the tail width be close to 0, while the head should maintain 70%–90% of the solid width to create a sense of "flow." Therefore, 80% was set as the most visually balanced value based on experience. Transparency and Gradient Control: In Three.js animations, a minimum tail width of 0.1 pixels and a completely transparent end color are the most commonly used "hidden shadow removal" parameters. These ensure that the trajectory does not leave obvious afterimages after it stops moving. Finally, the rendering engine draws on the screen based on the bound parameters and the coordinates in the displacement time chain. Specifically, it draws a series of connected line segments from the head (oldest coordinates) to the tail (newest coordinates). The color of the line segments smoothly transitions from the end color of the trail to the beginning color of the trail, and the width of the line segments also increases linearly from the end width of the trail to the beginning width of the trail, generating a dynamic trajectory display effect.

[0143] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A heating ventilation system operation data visualization processing system, characterized in that, The system comprises: The heat flux feature extraction module obtains the heat flux density value through the sensor network, filters out the noise by using the Kalman filtering algorithm, analyzes the node coordinates and the direction vector angle, generates the heat node graphical coordinate data set, and transmits it to the direction gradient difference calculation module; The direction gradient difference calculation module receives the heat node graphical coordinate data set, performs real-time difference operation on the direction vector angle values of adjacent nodes, identifies the direction change trend according to the difference sign, generates the adjacent node connection line rendering strategy, and transmits it to the fluctuation amplitude analysis module; The fluctuation amplitude analysis module receives the adjacent node connection line rendering strategy, calculates the heat flux density fluctuation amplitude value in time sequence, analyzes the fluctuation trend direction and the period change frequency by using the sliding window average method, generates the node path reconstruction instruction, and transmits it to the connection line attribute control module; The fluctuation amplitude analysis module comprises: The heat flux calculation submodule calls the adjacent node connection line rendering strategy, extracts the heat flux density change rate in the time sequence, compares the node transmission delay difference and the flux gradient difference, screens the time continuous difference sample set, calculates the heat flux density fluctuation characteristic value in the sliding window, and obtains the heat flux fluctuation amplitude result set; The fluctuation trend analysis submodule divides the time window based on the heat flux fluctuation amplitude result set, extracts the fluctuation mean value, calculates the mean value difference sign change sequence of adjacent windows, determines the reverse node according to the number of sign switches, and counts the period change frequency to obtain the fluctuation trend frequency spectrum vector; The node path instruction generation submodule calls the fluctuation trend frequency spectrum vector, screens the node sequence whose period change frequency is in the threshold interval, calculates the node connection line topology stability index and combines it with the frequency distribution characteristics for judgment, and generates the node path reconstruction instruction; The connection line attribute control module calls the node path reconstruction instruction, performs weighted operation on the direction gradient difference value and the color mapping coefficient to update the connection line color parameter, performs linear mapping calculation of the line width proportion coefficient and the gradient difference value to obtain the line width adjustment value, and outputs the visualization atlas rendering result to the dynamic trajectory optimization module; The connection line attribute control module comprises: The direction gradient difference value calculation submodule calls the node path reconstruction instruction, obtains the connection line direction vector data between nodes, extracts the node coordinate sequence, calculates the angle change according to the coordinate difference, performs difference and offset ratio judgment on the node direction change amount, arranges the direction offset proportion, and obtains the direction gradient difference value result; The color parameter weighted update submodule extracts the node color channel parameter based on the direction gradient difference value result, calculates the corresponding change amount of the gradient difference value and the color coefficient, summarizes the node color parameter difference, calculates the color mapping update parameter and maps it to the connection line color channel to form the connection line color parameter result; The line width mapping adjustment submodule extracts the connection line gradient difference value according to the connection line color parameter result, calculates the linear mapping amount of the proportion coefficient and the gradient difference value, analyzes the line width change gradient and superimposes the original line width sequence, and outputs the visualization atlas rendering result.

2. The HVAC system operational data visualization processing system of claim 1, wherein, The thermal node graph coordinate data set comprises a node space coordinate data set, a direction vector angle matrix and a heat flux density weighting coefficient set, the adjacent node connection line rendering strategy comprises a direction change sign set, a node connection line topology set and a rendering priority sequence set, the node path reconstruction instruction comprises a fluctuation trend direction parameter set, a periodical change frequency parameter set and a path coordinate adjustment matrix, and the visualization atlas rendering result comprises a color mapping parameter set, a line width proportion adjustment coefficient set and a layer display level set.

3. The HVAC system operational data visualization processing system of claim 1, wherein, The heat flux feature extraction module comprises: A heat flux acquisition sub-module acquires heat flux density values of sampling nodes through a sensor network, sorts sampling data according to node time sequence numbers, detects time stamp matching relationships and calculates flux difference intervals, compares time delays according to synchronization signals and reorganizes flux sequences, and generates node heat flux time sequence data sets; A noise filtering processing sub-module extracts adjacent sampling flux change rates based on the node heat flux time sequence data sets, calculates prediction and actual measurement residual sequences, performs state prediction and observation update by using a Kalman filtering algorithm, adjusts abnormal samples and updates index states, and obtains a smooth heat flux density sequence; A node feature calculation sub-module reads node space coordinates and direction vectors according to the smooth heat flux density sequence, calculates distance vectors and included angle differences, calculates gradient differences according to direction change rates and performs mapping, and obtains a thermal node graph coordinate data set.

4. The HVAC system operational data visualization processing system of claim 1, wherein, The direction gradient difference calculation module comprises: A spatial data analysis sub-module acquires the thermal node graph coordinate data set, topologically numbers node coordinate points, calculates Euclidean distance difference values based on node indexes, detects coordinate change rates according to time steps, calculates displacement direction components and relative distance intervals between adjacent nodes, and generates a node displacement vector set; A direction vector operation sub-module extracts adjacent node direction component coordinate differences based on the node displacement vector set, calculates unit direction vector angles, performs linear interpolation when the angle difference is less than an angle reference threshold, and generates a direction angle difference sequence; A difference trend identification sub-module calculates angle difference signs and detects change intervals according to the direction angle difference sequence, counts sign proportions, extracts direction fluctuation frequencies according to positive and negative sign trends, and generates an adjacent node connection line rendering strategy.

5. The HVAC system operational data visualization processing system of claim 4, wherein, The detection of the coordinate change rate according to the time step sequence comprises taking adjacent time intervals in the time step sequence as a reference to detect the change rate of the node coordinate point. The angle reference threshold is set based on the flow line deflection angle corresponding to the critical Reynolds number in the transition zone between laminar flow and turbulent flow in fluid mechanics.

6. The HVAC system operational data visualization processing system of claim 1, wherein, The color mapping update parameter is calculated by using the formula: ； wherein, representing a color mapping update parameter, representing an i-th node directional gradient value, representing a directional gradient reference value, representing an i-th node color mapping coefficient, m representing a total number of nodes.

7. The HVAC system operational data visualization processing system of claim 1, wherein, The dynamic trajectory optimization module receives the visualization atlas rendering result, calculates coordinate offsets based on fluctuation amplitude values and displacement coefficients, updates the coordinate position of the node in the atlas according to the fluctuation trend direction, performs real-time binding operation of the trajectory tailing parameter and the current coordinate, and generates a dynamic trajectory display effect; The dynamic trajectory display effect comprises a coordinate offset sequence set, a node motion trajectory data set and a tailing display control parameter set.

8. The HVAC system operational data visualization processing system of claim 7, wherein, The dynamic trajectory optimization module comprises: The rendering data analysis submodule obtains the visualization atlas rendering result, collects node fluctuation amplitude values and displacement coefficients, detects data integrity and eliminates abnormal items, performs proportional conversion according to the fluctuation amplitude values and the displacement coefficients, and generates node coordinate offset amounts; The displacement calculation submodule establishes a node initial coordinate set according to the node coordinate offset amounts, superimposes the node initial coordinate set and offset base amounts, extracts fluctuation trend direction parameters and adjusts coordinate offset directions through symbol matching, updates node real-time coordinate positions, and obtains node dynamic displacement amounts; The trajectory generation submodule calls the node dynamic displacement amounts to establish node coordinate update sequences, arranges the node coordinate update sequences into displacement time chains in chronological order, binds trajectory tailing parameters and node current coordinates, and generates a dynamic trajectory display effect.

Images