Operation state monitoring method of cold boiled water supply equipment

Abstract

The invention relates to the field of water supply equipment monitoring, in particular to an operation state monitoring method for cold boiled water supply equipment, which comprises the following steps: acquiring state data of multiple dimensions of the water supply equipment; classifying the historical water taking events, acquiring state data of the historical water taking events with the same category as the current water taking event, and calculating a reference value of the state data of each dimension in the historical water taking events; constructing an operation deviation degree vector by using a difference value between the state data at the current moment and a reference value; calculating the stability of the water supply equipment; the ratio of the norm of the operation deviation degree vector to the stability is used as a local sensitivity enhancement feature; and constructing an augmented state vector containing state data and local sensitivity enhancement features, and performing anomaly detection on the augmented state vector by using the detection model. According to the method, the detection model can sensitively capture early fault symptoms, so that accurate predictive maintenance is realized.

Description

Technical Field

[0001] This invention relates to the field of water supply equipment monitoring, and in particular to a method for monitoring the operating status of a boiled water supply equipment. Background Technology

[0002] As a crucial facility for providing safe and healthy drinking water, the continuous reliability of boiled water supply equipment is of paramount importance. To ensure stable operation and prevent unexpected downtime, a system of operational status monitoring is typically implemented to enable early warning of potential faults and predictive maintenance.

[0003] The Isolation Forest algorithm, due to its efficiency and the fact that it does not require pre-labeling fault samples, has been applied to anomaly detection in equipment operation data. This algorithm assesses the degree of anomaly by constructing multiple random trees and calculating the average path length required to completely isolate a data point. Its core idea is that outliers are easier to isolate due to their sparse and distinct characteristics.

[0004] However, traditional Isolation Forest algorithms face a significant technical problem when applied to systems with defined operating cycles, such as water supply equipment: essentially acting as a global anomaly detector, they exhibit a passivation effect, showing local insensitivity to indicators of slow performance degradation and gradual failures. Specifically, these early failures typically manifest as a slight but persistent shift in data points from a stable operating phase (such as insulation or standby mode) within its immediate vicinity of normally functioning data clusters. For the globally randomly partitioned Isolation Forest algorithm, is not significantly more difficult to isolate than normal points, making it difficult to effectively identify. For example, a small but continuous increase in pump runtime due to initial wear, or a slightly elevated outlet water temperature due to a minor decrease in cooling system efficiency, are often submerged in the global data distribution, causing the model to miss the optimal time for predictive maintenance and failing to effectively address equipment reliability and predictive maintenance issues. Summary of the Invention

[0005] To address the technical problem of poor monitoring performance caused by the insensitivity of the isolated forest algorithm to early gradual faults, this invention provides a method for monitoring the operating status of a boiled water supply system.

[0006] This invention provides a method for monitoring the operating status of a cooled boiled water supply device, employing the following technical solution: The system acquires multi-dimensional status data of the water supply equipment, including set temperature and outlet water temperature. Historical water intake events are categorized according to set temperature, and status data for historical water intake events of the same category as the current event are obtained. Reference values ​​for each dimension of status data in the historical water intake events are calculated. An operational deviation vector is constructed using the difference between the current status data and the reference values. The stability of the water supply equipment is calculated; the stability value is inversely proportional to the fluctuation of the difference between the outlet water temperature and the set temperature. The ratio of the norm of the operational deviation vector to the stability is used as a local sensitivity enhancement feature. An augmented state vector containing status data and local sensitivity enhancement features is constructed, and an anomaly detection model is used to perform anomaly detection on the augmented state vector.

[0007] By comparing the current state with recent historical events of the same category, a deviation vector is constructed, and local sensitivity enhancement features are calculated in conjunction with equipment stability. This amplifies weak signals indicating early, gradual failures, reducing their weight in unstable states and increasing their weight in stable states. Finally, by detecting the augmented state vector containing the enhanced features, the model's sensitivity and detection accuracy for local, gradual failures are improved, thus enabling early warning of potential equipment problems.

[0008] Preferably, the reference value is calculated as follows: the state data at the same moment in the historical water intake event is multiplied by the corresponding time weight coefficient to obtain the weighted state data. The time weight coefficient of the state data at the same moment decreases as the time interval with the current moment increases. The weighted state data at multiple moments are accumulated and then multiplied by the decay rate to obtain the reference value.

[0009] By employing time weighting and decay rate weighting calculations, higher weights are assigned to historical data more recent to the current moment. This method allows the reference benchmark to dynamically adapt to wear and aging of equipment due to normal use, reflecting the natural evolution trend of equipment condition. It avoids false alarms caused by using fixed, outdated benchmarks, making the benchmark for anomaly detection more timely and accurate.

[0010] Preferably, before calculating the reference value of the state data of each dimension in the historical water intake event, the method further includes: using the set temperature at the current moment to determine the category to which the current water intake event belongs; in the category to which it belongs, obtaining the historical water intake event closest to the current moment; obtaining the state data of each moment in the historical water intake event; and extracting the state data of the middle part.

[0011] A step of extracting data from intermediate periods of historical events was added before calculating the reference value. This is because the equipment's status data (such as temperature) will experience normal and significant fluctuations during the start-up and end phases of a water intake event; these fluctuations are not abnormal. By selecting intermediate data representing a stable operating phase as a reference, noise interference from these transient processes can be effectively filtered out, thus constructing a cleaner and more stable benchmark. This improves the accuracy of subsequent deviation calculations and avoids misjudging normal transient fluctuations as equipment malfunctions.

[0012] Preferably, the method for extracting the state data of the middle part is as follows: the state data of each moment in the historical water intake event is divided into multiple subsequences at equal intervals, the stability of each subsequence is detected by the ADF test algorithm, and multiple consecutive stable subsequences are merged to obtain the state data of the middle part.

[0013] The ADF test (unit root test) is used as a statistical tool to scientifically test the stationarity of data subsequences. It can objectively and automatically identify data intervals that are truly in a stable state, ensuring the reliability of the selected intermediate data. This provides a stronger guarantee for generating high-quality reference values ​​and further improves the robustness and accuracy of the monitoring method.

[0014] Preferably, the method for extracting the state data of the middle part is as follows: taking the state data of the middle moment of the historical water intake event as the midpoint, the state data of multiple moments are obtained along the left and right directions to obtain the state data of the middle part.

[0015] This method acquires data by expanding outwards from the midpoint of historical events, based on the reasonable assumption that the midpoint of an event best represents its stable operating state. In scenarios where computational efficiency is critical, this method can quickly and effectively filter out transient data during start-up and shutdown phases, while sacrificing a small amount of accuracy.

[0016] Preferably, the stability calculation method for the water supply equipment is as follows: taking the current time t as the endpoint, constructing a window along the historical direction, the expression for the stability of the cooled boiled water supply equipment at time t is: ; ; In the formula, This indicates the outlet water temperature of the boiled water supply equipment at time m. This represents the set temperature of the cooled boiled water supply equipment at time m, where m represents the index of the time sequence, and n+1 represents the number of times within the window. This represents the fluctuation factor of the outlet water temperature of the boiled water supply equipment at time m. This indicates the stability of the boiled water supply equipment at time t.

[0017] The stability of the equipment is quantified by calculating the variance of the difference between the outlet water temperature and the set temperature within a time window. This provides a key quantitative input for the subsequent construction of local sensitivity enhancement features, and can objectively evaluate the equipment's ability to maintain the target state within a specific time period, and can amplify small deviations during the stable period.

[0018] Preferably, before calculating the stability of the water supply equipment, the method further includes: using a preset time period after each water intake event as a buffer period; if the current time is within the buffer period, correcting the set temperature to obtain a corrected set temperature; replacing the set temperature with the corrected set temperature; and then calculating the fluctuation factor. The correction method is as follows: ; In the formula, The value m indicates the corrected set temperature of the cooled boiled water supply equipment. The value m indicates the set temperature of the cooled boiled water supply equipment. represents the set temperature of the cooled boiled water supply equipment during the last water intake event, p represents the total number of moments within the buffer period, and T represents the number of moments between moment m and the end of the buffer period.

[0019] By considering the thermal inertia effect between consecutive water intake events, when the interval between two water intakes is short and the set temperature changes, the outlet water temperature will deviate normally due to the residual heat from the previous event. This is not a equipment malfunction. By setting a buffer period and dynamically correcting the set temperature during this period, the actual physical cooling / heating process is simulated. This effectively avoids the system misjudging such predictable and normal temperature fluctuations as abnormalities, significantly reducing the false alarm rate and making stability calculations and anomaly detection results more accurate and reliable.

[0020] Preferably, before constructing the augmented state vector, the method further includes: normalizing the state data using a linear normalization algorithm.

[0021] Preferably, the status data includes: the set temperature of the cooled boiled water, the outlet temperature of the cooled boiled water, the cumulative working time of the solenoid valve, and the cumulative working time of the water supply pump.

[0022] In addition to standard temperature data, the cumulative operating time of solenoid valves and water pumps is also included. This provides richer equipment status information compared to monitoring only temperature.

[0023] Preferably, the detection model is the isolated forest model.

[0024] Isolation forest models offer advantages such as high computational efficiency and the elimination of the need for pre-labeled fault samples, making them highly suitable for unsupervised equipment anomaly detection. By constructing augmented state vectors, the insensitivity of traditional isolation forests to gradual faults is overcome.

[0025] The present invention has the following technical effects: By dynamically comparing the current state with stable data from similar historical events in recent times, minute operational deviations are quantified. These deviations are then amplified by weighting the operational stability of the equipment itself. This allows weak signals indicating slow performance degradation to be highlighted from background noise, enabling the detection model to sensitively capture early fault signs that are easily overlooked by traditional methods. This, in turn, enables accurate and timely predictive maintenance. Attached Figure Description

[0026] Figure 1 This is a flowchart of a method for monitoring the operating status of a cooled boiled water supply device according to the present invention. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] This invention discloses a method for monitoring the operational status of a cooled boiled water supply system, referring to... Figure 1 The process includes the following steps, as detailed below: S1: Obtain the status data of the boiled water supply equipment.

[0029] Multiple sensors inside the cooled boiled water supply equipment collect multi-dimensional status data in real time during operation. This data includes the set temperature of the cooled boiled water, the outlet temperature of the cooled boiled water, the cumulative operating time of the solenoid valve, and the cumulative operating time of the water supply pump. A linear normalization algorithm is used to normalize the status data to eliminate the influence of dimensions. A state vector is constructed using the status data collected at the same time. The state vector at time t is denoted as... : ; in, These represent the set temperature of the cooled boiled water, the outlet temperature of the cooled boiled water, the cumulative working time of the solenoid valve, and the cumulative working time of the water supply pump at time t, respectively. They can also be understood as... These represent the state data of the cooled boiled water supply equipment in the first, second, third, and fourth dimensions at time t.

[0030] S2: Construct the instantaneous running deviation vector.

[0031] Cooled boiled water supply equipment is usually set with multiple cooled boiled water temperatures, which correspond to multiple set temperatures, such as 40 degrees, 50 degrees and 55 degrees, for users to choose from. Each time a user takes water, they first select a set temperature and then take water. The time from selecting the set temperature to finishing taking water is defined as a water taking event.

[0032] To capture the minute drift of the device state relative to its recent normal performance, an instantaneous operational deviation vector needs to be constructed to quantify the difference between the current device state and its immediate neighboring, dynamically changing local steady state.

[0033] Historical water extraction events are categorized according to set temperatures. For example, water extraction events with a set temperature of 40 degrees Celsius are grouped into one category, water extraction events with a set temperature of 50 degrees Celsius are grouped into another category, and water extraction events with a set temperature of 55 degrees Celsius are grouped into a third category. Within each category, water extraction events are arranged in chronological order.

[0034] The current water intake event is classified according to the set temperature at the current time t. Within the classified category, the historical water intake event closest to the current time t is obtained. The state data of each time point in the historical water intake event is obtained, the state data of the middle part is extracted, and the reference value of the state data of each dimension in the historical water intake event is calculated.

[0035] In one embodiment, the method for obtaining the state data of the intermediate part is as follows: the state data of each moment in the historical water intake event is divided into multiple subsequences at equal intervals, the stability of each subsequence is detected by the ADF test algorithm, and multiple consecutive stable subsequences are merged to obtain the state data of the intermediate part.

[0036] In another embodiment, the method for obtaining the state data of the middle part is as follows: taking the state data of the middle moment of the historical water intake event as the midpoint, the state data of multiple moments are obtained along the left and right directions to obtain the state data of the middle part.

[0037] For example, if the set temperature for this water intake event is 40 degrees, and the state data of each moment in the previous water intake event with a set temperature of 40 degrees is obtained, and the number of moments is 50, then the state data of moments 21 to 30 are extracted.

[0038] The method for calculating the reference value is as follows: ; ; In the formula, This represents the reference value of the nth dimension of the state data in the historical water intake event closest to the current time t. This represents the state data of the nth dimension at time k in a historical water extraction event. This represents the weighting coefficient, and tk represents the time length between time t and time k, in seconds. This represents the adjustment coefficient, used to adjust the sensitivity of the weighting coefficient to the length of time. For example, the adjustment coefficient is 0.8. The attenuation rate of the water supply equipment is represented by the following calculation method: calculate the time interval between the current water intake event and the most recent historical water intake event of the same category, and multiply the time interval by the basic attenuation rate as the attenuation rate of the water supply equipment.

[0039] During the operation of a boil-and-cool water supply system, normal aging processes occur over time, leading to deviations in its status data. For example, aging of the heater causes a slight deviation in the outlet water temperature, potentially resulting in a temperature lower than the factory-set temperature. Similarly, scale buildup reduces the container capacity, consequently decreasing the cumulative operating time of the solenoid valve. The baseline decay rate represents the normal decrease in status parameters per unit time, and its value is determined based on historical water intake data or experimental results.

[0040] For example, in a water intake event at 40 degrees Celsius, the attenuation rate was 0.99. The value is 39 degrees, then The value is 39 × 0.99, which represents a normal value.

[0041] The reference value reflects the benchmark for the state data in the corresponding dimension. As the water supply equipment is used, internal parts undergo normal wear and tear and aging, causing normal changes in the state data. Therefore, when calculating the reference value for each dimension of state data, the state data most recent to the current moment is given a higher weight, ensuring that the benchmark is updated over time and avoiding misjudgments caused by relying on a fixed benchmark. In essence, each dimension of state data corresponds to a reference value.

[0042] The operational deviation vector is constructed using the difference between the state data at time t and the reference value. The expression for the operational deviation vector is: ; in, This represents the operational deviation vector of the state data at time t. This represents the difference between the set temperature of cooled boiled water at time t and the reference value of the set temperature in historical events. This represents the difference between the outlet temperature of cooled boiled water at time t and the reference value of the outlet temperature in historical events. This represents the difference between the cumulative operating time of the solenoid valve at time t and the reference value of the cumulative operating time of the solenoid valve in historical events. This represents the difference between the cumulative operating time of the water supply pump at time t and the reference value of the cumulative operating time of the water supply pump in historical events.

[0043] Each component of the operational deviation vector represents the degree and direction of deviation of the corresponding operational parameter from its recent average performance at the current moment. A non-zero operational deviation vector indicates that the device state has drifted.

[0044] S3: Calculate the stability of the cooled boiled water supply equipment.

[0045] The stability of a cooled boiled water supply system is typically measured by the difference between the outlet water temperature and the set temperature. Under normal circumstances, the outlet water temperature should fluctuate slightly relative to the set temperature. The closer the outlet water temperature is to the set temperature, the more stable the system's operation. A buffer period of two minutes is defined after each water intake event. During this period, a normal deviation from the set temperature is expected. For example, if the set temperature for the previous water intake event was 55 degrees Celsius, and the set temperature for this event is 40 degrees Celsius, and the event occurred recently, even if the system is functioning normally, the outlet water temperature will still be 40 degrees Celsius higher than the set temperature due to the higher temperature in the outlet pipes. Therefore, based on this principle, the stability of a cooled boiled water supply system can be calculated as follows: If the current time t is not within the buffer period, a window is constructed along the historical direction with the current time t as the endpoint. The size of the window is manually selected according to the actual situation. For example, the length of the window is 20, meaning the number of times within the window is 20. If the window intersects with the buffer period, the data of the intersecting times are discarded. The expression for the stability of the boiled water supply equipment at time t is:

[0046]

[0047] In the formula, This indicates the outlet water temperature of the boiled water supply equipment at time m. This represents the set temperature of the cooled boiled water supply equipment at time m, where m represents the index of the time sequence, and n+1 represents the number of times within the window. This represents the fluctuation factor of the outlet water temperature of the boiled water supply equipment at time m. This indicates the stability of the boiled water supply equipment at time t.

[0048] This reflects the degree to which the outlet water temperature of the boiled water supply equipment deviates from the set temperature at time m. The larger the value, the greater the deviation of the outlet water temperature from the set temperature. At this time, the lower the stability, the greater the possibility that the boiled water supply equipment is in an abnormal state.

[0049] If the current time t is within the buffer period, then the stability expression of the cooled boiled water supply equipment at time t is: ; ; ; In the formula, This represents the fluctuation factor of the outlet water temperature of the boiled water supply equipment at time m. This indicates the outlet water temperature of the boiled water supply equipment at time m. The value m indicates the corrected set temperature of the cooled boiled water supply equipment. The value m indicates the set temperature of the cooled boiled water supply equipment. This represents the set temperature of the cooled boiled water supply equipment during the last water intake event, m represents the index of the time sequence, p represents the total number of times within the buffer period, and T represents the number of times between time m and the end of the buffer period. This indicates the stability of the boiled water supply equipment at time t.

[0050] For example, the last water collection ended at The buffer period corresponds to the time period of [ , The current time t is within the buffer period. , If the current time t < ], then the current time t < p represents to The number of time intervals between time m and time T, where T represents time m and time m. The number of time points in between, i.e., the end of the buffer period is time point. .

[0051] If the time interval between two water intake events is short, the set temperature of the previous water intake event will have a certain impact on the outlet water temperature of the current water intake event. If the current time t is relatively far from the previous water intake event (a larger t value), the set temperature of the current boiled water supply equipment will have a greater weight; if the current time t is relatively close to the previous water intake event (a smaller t value), the set temperature of the previous boiled water supply equipment will have a greater weight. By correcting the set temperature, a reference baseline for the outlet water temperature is simulated. Ideally, within the buffer period, Should revolve around It will fluctuate slightly.

[0052] After calculating the fluctuation factor, the stability of the cooled boiled water supply equipment is then calculated using the fluctuation factor. Therefore, by correcting the set temperature, the impact of normal temperature fluctuations on the calculated stability is reduced.

[0053] S4: Calculate the local sensitivity enhancement feature.

[0054] In order for the Isolation Forest algorithm to detect local anomalies that occur with small deviations during the stable phase, it is necessary to calculate the augmented features at time t, expressed as: ; in, For the local sensitivity enhancement feature at time t; Let be the Euclidean norm of the deviation vector of the state data at time t, representing the overall magnitude of the current state's deviation from its local baseline; This indicates the stability of the boiled water supply equipment at time t.

[0055] The larger the value, the greater the deviation in the current state data, and the higher the probability that the boiled water supply equipment will malfunction in the future. This indicates the stability of the boiled water supply equipment. The smaller the value, the more unstable the boiled water supply equipment is. Therefore, the enhanced local sensitivity characteristic can comprehensively reflect the health status of the boiled water supply equipment's operating condition.

[0056] S5: Construct an augmented state vector to monitor the operating status of the boiled water supply equipment.

[0057] ; In the formula, This represents the augmented state vector at time t. This represents the state vector at time t. This represents the local sensitivity enhancement feature at time t. The augmented state vector retains the original state information while also carrying additional local context-sensitive information about state deviations. A sequence dataset consisting of augmented state vectors from multiple time points is used as input to train the Isolation Forest model, resulting in a detection model. During operation, the augmented state vectors at corresponding time points are acquired in real-time, and their anomaly scores are calculated using the detection model to monitor the operating status of the boiled water supply equipment.

[0058] Because the augmented state vector includes local sensitivity enhancement features, for a data point that deviates slightly from a state with poor stability, its... The value will be amplified ( The value is low), making this point stand out from other normal points in the newly added dimension (its value is low). The values ​​are generally lower, which is significantly different. Therefore, when performing random splitting in an isolated forest, this point will be considered because... Its dimensional uniqueness makes it easily isolated, resulting in a very low path length and a very high anomaly score.

[0059] Finally, by setting an anomaly score threshold θ, the operating status of the equipment is judged. When the anomaly score is greater than the threshold θ, it can be determined that the equipment has shown signs of early, gradual failure, and corresponding early warnings or maintenance work orders are triggered, thereby achieving highly sensitive monitoring of local anomalies.

[0060] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for monitoring the operating status of a cooled boiled water supply device, characterized in that, Including the following steps: The system acquires multi-dimensional status data of the water supply equipment, including set temperature and outlet water temperature. Historical water intake events are categorized according to set temperature, and status data of historical water intake events of the same category as the current water intake event is obtained. Reference values ​​for each dimension of status data in the historical water intake events are calculated. An operational deviation vector is constructed using the difference between the current status data and the reference values. The stability of the water supply equipment is calculated; the stability value is inversely proportional to the fluctuation of the difference between the outlet water temperature and the set temperature. The ratio of the norm of the operational deviation vector to the stability value is used as a local sensitivity enhancement feature. An augmented state vector containing state data and local sensitivity enhancement features is constructed, and an anomaly detection model is used to perform anomaly detection on the augmented state vector.

2. The method for monitoring the operating status of a cooled boiled water supply device according to claim 1, characterized in that, The reference value is calculated as follows: the state data at the same moment in the historical water intake event is multiplied by the corresponding time weight coefficient to obtain the weighted state data. The time weight coefficient of the state data at the same moment decreases as the time interval with the current moment increases. The weighted state data at multiple moments are accumulated and then multiplied by the decay rate to obtain the reference value.

3. The method for monitoring the operating status of a cooled boiled water supply device according to claim 1, characterized in that, Before calculating the reference value of the status data of each dimension in the historical water intake event, the process also includes: using the set temperature at the current moment to determine the category to which the current water intake event belongs; in the category to which it belongs, obtaining the historical water intake event closest to the current moment; obtaining the status data of each moment in the historical water intake event; and extracting the status data of the middle part.

4. The method for monitoring the operating status of a cooled boiled water supply device according to claim 3, characterized in that, The method for extracting the state data of the middle part is as follows: the state data of each moment in the historical water intake event is divided into multiple subsequences at equal intervals, the stability of each subsequence is detected by the ADF test algorithm, and multiple consecutive stable subsequences are merged to obtain the state data of the middle part.

5. The method for monitoring the operating status of a cooled boiled water supply device according to claim 3, characterized in that, The method for extracting the state data of the middle part is as follows: take the state data of the middle moment of the historical water intake event as the midpoint, and obtain the state data of multiple moments along the left and right directions to obtain the state data of the middle part.

6. The method for monitoring the operating status of a cooled boiled water supply device according to claim 1, characterized in that, The stability calculation method for water supply equipment is as follows: Using the current time t as the endpoint, a window is constructed along the historical direction. The expression for the stability of the cooled boiled water supply equipment at time t is: ； ； In the formula, This indicates the outlet water temperature of the boiled water supply equipment at time m. This represents the set temperature of the cooled boiled water supply equipment at time m, where m represents the index of the time sequence, and n+1 represents the number of times within the window. This represents the fluctuation factor of the outlet water temperature of the boiled water supply equipment at time m. This indicates the stability of the boiled water supply equipment at time t.

7. The method for monitoring the operating status of a cooled boiled water supply device according to claim 6, characterized in that, Before calculating the stability of the water supply equipment, the following steps are taken: A preset time period after each water intake event is used as a buffer period. If the current time falls within this buffer period, the set temperature is corrected to obtain a corrected set temperature. This corrected set temperature replaces the original set temperature. Then, the fluctuation factor is calculated. The correction method is as follows: ； In the formula, The value m indicates the corrected set temperature of the cooled boiled water supply equipment. The value m indicates the set temperature of the cooled boiled water supply equipment. represents the set temperature of the cooled boiled water supply equipment during the last water intake event, p represents the total number of moments within the buffer period, and T represents the number of moments between moment m and the end of the buffer period.

8. The method for monitoring the operating status of a cooled boiled water supply device according to claim 1, characterized in that, Before constructing the augmented state vector, the process also includes normalizing the state data using a linear normalization algorithm.

9. The method for monitoring the operating status of a cooled boiled water supply device according to claim 1, characterized in that, Status data includes: the set temperature of boiled water, the outlet temperature of boiled water, the cumulative working time of the solenoid valve, and the cumulative working time of the water supply pump.

10. The method for monitoring the operating status of a cooled boiled water supply device according to claim 1, characterized in that, The detection model is the isolated forest model.

Images