A multi-site predictive filter end-of-life time indicating system and method

Abstract

The application provides a kind of multi-site prediction filter failure time indication system and method, belong to filter service life monitoring field.Solve the problems of poor reliability of existing prediction device, low prediction accuracy in complex environment and adverse effects on filter filtering performance in prediction process.The prediction device mentioned in the application can accurately predict the filter failure time under complex conditions, providing a margin time for replacing a new filter.The prediction device has a simple structure and high reliability, and is suitable for use in various occasions.It operates independently without affecting the filtering capacity and structure of the filter during operation.Meanwhile, the prediction device can automatically adjust the prediction node according to the filter specifications for accurate prediction.

Description

Technical Field

[0001] This invention belongs to the field of filter life monitoring, and specifically relates to an indication system and method for predicting filter failure time at multiple locations. Background Technology

[0002] With the rapid development of modern industrial technology, and to meet the demands of development, toxic gases such as toluene, diethyl ether, hydrogen cyanide, dichloromethane, halogenated hydrocarbons, ammonia, alcohols, esters, and ketones are widely used in chemical production processes. The extensive use of hazardous chemicals greatly increases the probability of direct contact between toxic gases and personnel. To ensure personnel safety, air purification is necessary in environments with potential toxic gas leaks. In the purification process, various adsorbents are filled inside filters to remove toxic gases. However, due to the complex and variable external environment, the effective time of filters obtained under experimental conditions differs significantly from their effective time in actual application. Knowing the effective time of filters in advance facilitates timely replacement of filter cartridges, ensuring personnel safety. Therefore, there is an urgent need for a device that can avoid complex environmental factors, is widely applicable in various situations, and accurately predicts filter failure time.

[0003] The invention patent with application number CN 113237682 discloses a filter canister protection time detection device and its usage method. The method opens sampling ports at different heights of the filter canister and leads out a branch at each sampling port. Each branch is equipped with a sensor and an alarm module. The degree of use of the filter canister is determined by judging the concentration of gas in the branch. This method directly leads out the branch on the canister, which can easily disrupt the gas flow pattern in the canister and thus affect the adsorption efficiency of the adsorbent. In addition, the sampling ports on the canister are fixed, which means that the prediction can only be made for fixed time points.

[0004] Patent application number CN 209446430 proposes a method for predicting the lifespan of filters used in civil defense projects. This method involves setting up a branch at the filter outlet and analyzing the gas in the branch to determine whether the filter has failed. This method can accurately determine when the filter will fail, but it cannot predict the remaining effective time of the filter in advance to allow sufficient time for filter replacement. For some highly toxic gases, even a small leak can cause serious harm. Therefore, real-time dynamic prediction of the filter's lifespan is crucial.

[0005] The purpose of this invention is to provide a device and method for predicting the remaining effective time of a filter, addressing the problems of poor reliability, low prediction accuracy in complex environments, and adverse effects on filter performance during the prediction process in existing prediction devices. The prediction device mentioned in this invention can accurately predict filter failure time under complex conditions, providing sufficient time for filter replacement. This prediction device has a simple structure, high reliability, and is suitable for various applications. It operates independently of the filter, without adversely affecting the filter's filtration capacity or structure. Furthermore, the prediction device can automatically adjust the prediction interval according to the filter specifications for accurate prediction. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention proposes a prediction device and method with wide applicability, high reliability, and automatic adjustment of prediction nodes. By controlling the opening and closing of valves, the number and length of adsorption columns connected to the prediction device are changed, thereby predicting the remaining effective time of the filter at different time points.

[0007] The technical solution of the present invention:

[0008] A multi-site predictive system for filter failure time includes a control system, a concentration detection system, a prediction system, and a flow regulation system. The prediction system comprises n prediction units, each consisting of m adsorption columns of equal or unequal length. All adsorption columns in all prediction units must have the same diameter. Different prediction units may have the same or different number of adsorption columns, and the lengths of the adsorption columns in different prediction units may be equal or unequal. In each prediction unit, the adsorption columns are connected in parallel, and each adsorption column is equipped with a valve. Prediction units are connected in series. Except for the last prediction unit, the main path at the rear end of each prediction unit is connected to the inlet of the concentration detection system, and a valve is installed between the main path at the rear end of each prediction unit and the concentration detection system. The last prediction unit is directly connected to the inlet of the concentration detection system. By adjusting the opening and closing of the internal valves, adsorption columns of different lengths and numbers in each prediction unit are connected, thereby achieving automatic switching of prediction nodes and predicting the remaining time of the filter at different time points. The outlet of the concentration detection system is connected to the pump inlet, and the pump outlet is connected to the polluted environment; the control system is connected to the concentration detection system, the prediction system, the flow regulation system, and the pump; the flow regulation system is connected to the inlet of the first prediction unit; the adsorption column and the filter are filled with the same adsorbent and work synchronously under the same operating conditions to avoid the influence of complex external conditions on the indicating system.

[0009] Furthermore, n is 4 or 5, m is 2-4, and the same operating conditions for the adsorption column and filter refer to the same inlet gas type, inlet concentration, empty bed gas velocity, and operating temperature.

[0010] Furthermore, the flow regulation system is used to regulate the flow rate entering the indicator system, ensuring that the indicator system and the filter work synchronously.

[0011] The control system is used to control the working status of each electric component and display it on the panel, as well as to store and process the predicted data.

[0012] The concentration detection system has a built-in concentration sensor or an external concentration detection device (such as chromatography, mass spectrometry, infrared spectroscopy, etc.).

[0013] The method for predicting filter failure time at multiple sites using the aforementioned multi-site predictive filter failure time indication system requires prediction of the remaining time of the filter at different time points for different applications. Therefore, by connecting adsorption columns of different lengths and numbers built into the indication system, the prediction of the remaining effective time of the filter at different time points is achieved, as detailed below:

[0014] (1) Open the corresponding valve in the prediction system according to the filter size and the predicted time node, and connect the adsorption columns of different lengths and numbers.

[0015] (2) Obtain the filter operating conditions and adjust the inlet flow of the prediction system through the flow regulation system to ensure that the prediction system and the filter operating conditions are the same; the filter operating conditions refer to the inlet gas type, inlet concentration, empty bed gas velocity and operating temperature.

[0016] (3) The filter and prediction system start working simultaneously:

[0017] Based on the filter operating conditions obtained in step (2), when the relationship between the packing length and the breakthrough time in the adsorption column is known for the filter under this operating condition, the remaining effective time of the filter can be calculated directly by the packing length in each prediction unit.

[0018] Based on the filter operating conditions obtained in step (2), when the relationship between the packing length and breakthrough time in the adsorption column is unknown for the filter under this operating condition, then, on the basis of ensuring that the total packing length of the prediction unit is equal to the filter packing length, and according to the principle that the ratio of the breakthrough time corresponding to the same packing length to the total breakthrough time under different conditions remains unchanged, combined with the relationship between the packing length and breakthrough time in a certain adsorption column obtained in advance under experimental conditions, the remaining failure time of the filter under any condition is calculated by the following formula:

[0019]

[0020] Among them, t i * t represents the predicted breakthrough time of the unit under unknown conditions. b *Represents different values ​​of t under unknown conditions i * The corresponding remaining valid time; t i The value represents the predicted unit breakthrough time under experimental conditions, and T represents the filter breakthrough time under experimental conditions.

[0021] Furthermore, in step (1), the filter gauge dimensions include the filter length and cross-sectional area.

[0022] Furthermore, in step (1), the principle for the length and number of adsorption columns connected is to ensure that the total packing length or equivalent length of the adsorption columns is equal to the length of the filter packing.

[0023] Furthermore, in step (1), the methods for selecting the length and number of adsorption columns include the following two categories:

[0024] Method 1: Select each prediction unit to connect only one adsorption column, and the packing lengths of the connected adsorption columns can be the same or different; then the relationship between the adsorption column and the prediction time is linear:

[0025] y = ax + b(1)

[0026] Where y represents the breakthrough time and x represents the length of the adsorption column packing.

[0027] Method 2: When selecting at least one prediction unit that connects two or more adsorption columns, the packing length of all connected adsorption columns within the same prediction unit must be equal. The packing lengths of adsorption columns in different prediction units may be the same or different. Multiple adsorption columns are treated as a single adsorption column for equivalent prediction.

[0028] y=a(x+ΣZ)+b(2).

[0029] Where Z is the equivalent length of the adsorption column packing material to x, except for x; and x is the packing material length of any adsorption column within each prediction unit.

[0030] Further, in step (2), the method of adjusting the inlet flow rate of the prediction system through the flow regulation system is as follows: when the diameter of the adsorption column in the indicator system is equal to the diameter of the filter, the flow rates of the indicator system and the filter are equal. If the two diameters are not equal, let the filter inlet flow rate be Q, the filter diameter be D, the adsorption column diameter be d, and the indicator system inlet flow rate be Q. y Then the flow regulation system must ensure that the flow at the inlet of the indicator system satisfies the following relationship:

[0031]

[0032] Furthermore, in addition to predicting filter failure time, the indicator system can also monitor the external environment in real time, including real-time monitoring of toxic gas concentration, temperature, and humidity.

[0033] Furthermore, the concentration detection system, used to detect the breakthrough concentration of the prediction unit, can have a corresponding concentration sensor built into the device, or it can be connected to an external detection instrument through a reserved interface.

[0034] Furthermore, the control system consists of a PLC and a control panel, which is mainly used to adjust the opening and closing of each valve and the speed of the air pump. The control panel is used to display the operating parameters of each component and the remaining effective time of the filter.

[0035] The beneficial effects of this invention are as follows: The multi-site predictive filter failure time indication system mentioned in this invention can predict the remaining effective time of the filter at different time points by controlling the number of adsorption columns and the length of the packing material in the indicator system, making it adaptable to more situations. At the same time, the prediction indicator system operates in the same environment as the filter, and through the flow regulation system, it ensures that the flow rate, temperature, and gas properties of the prediction indicator system and the filter are equal, so as to achieve the purpose of working synchronously with the filter and avoid the influence of complex environmental factors on the prediction accuracy. In addition, under unknown environmental conditions, when each prediction unit fails, the prediction indicator system can still predict the failure time of the filter based on the relationship between the failure time and the packing length and failure time under any conditions, thus improving the reliability of the filter in field operations.

[0036] Instruction manual illustrations

[0037] Figure 1 Overall layout diagram of the prediction device.

[0038] Figure 2 Example 1: Prediction principle diagram.

[0039] Figure 3 Example 2: Prediction principle diagram.

[0040] In the diagram: 1-12 represent the first to twelfth adsorption columns; 13-27 represent the first to fifteenth valves; 28 represents the concentration detector; 29 represents the pump; 30-41 represent the thirteenth to twenty-fourth adsorption columns; 42-56 represent the sixteenth to thirtieth valves. Detailed Implementation

[0041] The following embodiments further illustrate the multi-site predictive filter failure time indication system proposed in this invention.

[0042] In the embodiments, the breakthrough concentration refers to the outlet concentration reaching 5% of the inlet concentration, the breakthrough time refers to the time from when the toxic gas enters the prediction unit to when the outlet concentration reaches the breakthrough concentration, and the effective time of the filter is also the breakthrough time.

[0043] Example 1: This example mainly verifies the accuracy of the prediction of the remaining effective time by different prediction nodes when adsorption columns with different packing lengths are connected.

[0044] (1) The apparatus of this embodiment:

[0045] The schematic diagram of the device in this embodiment is as follows: Figure 2 As shown, the prediction system consists of four prediction units: the first prediction unit, the second prediction unit, the third prediction unit, and the fourth prediction unit. Units 1-12 are the first to twelfth adsorption columns comprising the prediction unit; units 13-27 are the first to fifteenth valves; unit 28 is the concentration detector; and unit 29 is the pump. The packing length of each adsorption column in this prediction system is shown in Table 1.

[0046] Table 1

[0047]

[0048] In Table 1, “1, 2, 3, 4” represent different prediction units, and “L1, L2, L3” represent different adsorption columns within the same prediction unit. For example, the second column of the third column indicates that the packing length of the second adsorption column in the third prediction unit is 32 mm.

[0049] (2) Experiment based on the device of this embodiment

[0050] In this embodiment, NO is used as the toxic gas, with an inlet flow rate of 94.2 ml / min and an inlet concentration of 0.4 mol / m³. 3 .exist Figure 2In the schematic diagram of the device, adsorption columns with packing lengths of 80mm, 60mm, 40mm, and 20mm are connected. This involves opening valves 13 (first), 16 (fourth), 19 (seventh), 22 (tenth), 25 (thirteenth), 26 (fourteenth), and 27 (fifteenth) to conduct a prediction experiment. Toxic gas enters the first prediction unit at a certain flow rate. After a breakthrough in the first prediction unit, the toxic gas reaches the concentration detector through a branch between the first and second prediction units. The concentration detector detects the toxic gas concentration. When the breakthrough concentration is reached, the control system controls valve 25 to close. After a breakthrough in the second prediction unit, the toxic gas reaches the concentration detector through a branch between the second and third prediction units. When the toxic gas concentration reaches the breakthrough concentration, the control system controls valve 26 to close. After a breakthrough in the third prediction unit, the toxic gas reaches the concentration detector through a branch between the third and fourth prediction units. When the toxic gas concentration reaches the breakthrough concentration, the control system controls valve 27 to close. After a breakthrough in the fourth prediction unit, the prediction indication system terminates the prediction.

[0051] Under the above conditions, a prediction experiment was conducted on the prediction indication system. During the prediction process, the breakthrough time of each prediction unit was recorded, as shown in Table 2.

[0052] Table 2

[0053] Prediction Unit 1 2 3 4 Breakthrough Time 12 27 37 42

[0054] This predictive indication system can predict filter failure time for filters with a filling length of 200mm. Under the same conditions, the filter failure time is 40s, while the predictive indication system's failure time is 42s, with a prediction error of 5%. Without considering experimental errors, the method for predicting filter failure time proposed in this invention is reliable. When the first predictive unit fails, it indicates that the filter has been working for 12s, and the remaining effective time is 30s. When the second predictive unit fails, it indicates that the filter has been working for 27s, and the remaining effective time is 15s.

[0055] As shown in Table 2, the prediction indication system predicts the filter at a time ratio of 12:15:10:5. Depending on the application, predictions are required for different time points. By switching adsorption columns with different packing lengths, predictions can be made for different time points.

[0056] Based on the data in Tables 1 and 2, the relationship between the adsorption column packing length and the breakthrough time in this embodiment is as follows:

[0057] y = 0.25x - 7.94(5)

[0058] y: Breakthrough time; x: Adsorption column packing length.

[0059] (3) Verification test

[0060] For example, to predict the case with a time ratio of 10:5:4:2, it is necessary to calculate the adsorption column packing lengths of each prediction unit to be connected based on the time ratio and Equation (5), which are 112mm, 40mm, 32mm, and 16mm respectively. Figure 2 In the schematic diagram of the device shown, valves 14 (second valve), 17 (fifth valve), 20 (eighth valve), 23 (eleventh valve), 25 (thirteenth valve), 26 (fourteenth valve), and 27 (fifteenth valve) are opened and connected to an adsorption column of the aforementioned packing length for a prediction experiment. The breakthrough time of each prediction unit is shown in Table 3. The breakthrough time of the prediction indication system is 42s, and the prediction error is 5%.

[0061] Table 3

[0062] Prediction Unit 1 2 3 4 Breakthrough Time 20 30 38 42

[0063] Example 2: This example mainly verifies the accuracy of the remaining effective time predicted by different prediction nodes when the length and quantity of packing material connected to the adsorption column are changed simultaneously.

[0064] (1) The apparatus of this embodiment:

[0065] The schematic diagram of the device in this embodiment is as follows: Figure 3 As shown, the prediction system consists of four prediction units: the first, second, third, and fourth prediction units. Units 30-41 comprise the thirteenth to twenty-fourth adsorption columns, units 42-56 comprise the sixteenth to thirtieth valves, unit 28 is the concentration detector, and unit 29 is the pump. The filter length corresponding to this prediction indication system is 600 mm. The packing lengths of each adsorption column in this prediction device are shown in Table 4.

[0066] Table 4

[0067]

[0068] (2) Experiment based on the device of this embodiment

[0069] In this embodiment, NO is used as a toxic gas, and the inlet flow rate is controlled at 94.2 ml / min, with an inlet concentration of 0.4 mol / m³. 3The basic operating procedure of the device in this embodiment is as follows: After the first prediction unit fails, the toxic gas reaches the concentration detector through the branch between the first and second prediction units. The concentration detector detects the concentration of the toxic gas. When the failure concentration is reached, the control system controls the twenty-eighth valve 54 to close, and the toxic gas flows to the second prediction unit. After the second prediction unit fails, the toxic gas reaches the concentration detector through the branch between the second and third prediction units. When the concentration of the toxic gas reaches the failure concentration, the control system controls the branch valve twenty-ninth valve 55 to close, and the toxic gas flows to the third prediction unit. After the third prediction unit fails, the toxic gas reaches the concentration detector through the branch between the third and fourth prediction units. When the concentration of the toxic gas reaches the failure concentration, the control system controls the branch valve thirtieth valve 56 to close, and the toxic gas flows to the fourth prediction unit. After the fourth prediction unit fails, the prediction indication system ends the prediction.

[0070] (3) Experiments were conducted with each prediction unit connected to only one adsorption column.

[0071] Under the above conditions, each prediction unit of the prediction system is connected to one adsorption column during the prediction process. Specifically, the first, second, third, and fourth prediction units open valves 42 (sixteenth), 45 (nineteenth), 48 (twenty-second), 51 (twenty-fifth), 54 (twenty-eighth), 55 (twenty-ninth), and 56, respectively. The breakthrough time for each prediction unit, as shown in Table 5, is obtained from the prediction experiments.

[0072] Table 5

[0073] Prediction Unit 1 2 3 4 Breakthrough Time 16 42.5 68.7 95.4

[0074] The data in Tables 4 and 5 show that the length of the adsorption column packing material and the breakthrough time are linearly related, as shown in Equation (6).

[0075] y = 0.2546x - 6.7769 (6).

[0076] (4) Experiments were conducted with each prediction unit connected to two adsorption columns.

[0077] Under the above conditions, in such Figure 3 In the prediction system shown, each prediction unit is connected to two adsorption columns. The packing length of the adsorption columns connected in the same prediction unit is equal. That is, valves 42, 43, 45, 46, 48, 49, 51, 52, 54, 55, and 56 are opened. The breakthrough time of each prediction unit is recorded through prediction experiments, as shown in Table 6.

[0078] Table 6

[0079] Prediction Unit 1 2 3 4 Breakthrough Time 32.8 81.8 138 195.6 equivalent length 55 93 120 127

[0080] The two adsorption columns in each prediction unit are equivalent to one adsorption column. The equivalent length of each prediction unit is shown in Table 6. The equivalent relationship (7) is obtained from Table 6:

[0081] Z1 = 24.3n + 38(7)

[0082] Where n: number of prediction units

[0083] The above experiments yielded the equivalent relationship between the prediction unit and the breakthrough time after connecting different numbers of adsorption columns to the prediction indication system, as shown in Equation 8.

[0084]

[0085] According to Equation 8, predictions can be made for existing filters.

[0086] (5) Experiment 1 with any number of adsorption columns selected as the prediction unit

[0087] In such Figure 3 In the prediction system shown, valves 42 (16th), 43 (17th), 47 (21st), 48 (22nd), 49 (23rd), 51 (25th), 54 (28th), 55 (29th), and 56 (30th) are opened. Using Equation 8, the equivalent length and breakthrough time of each prediction unit after opening the above valves are calculated, as shown in Table 7.

[0088] Table 7

[0089] 1 2 3 4 equivalent length 162 125 211 100 Theoretical breakthrough time 34 66 120 145

[0090] Simultaneously, NO is used as a toxic gas, with an inlet flow rate of 94.2 ml / min and an inlet concentration of 0.4 mol / m³. 3 The breakthrough time of each predicted unit was obtained from the experiment, as shown in Table 8.

[0091] Table 8

[0092] Prediction Unit 1 2 3 4 Breakthrough Time 31 68 118 150

[0093] Under the above conditions, the predicted breakthrough time of the indicator system is 150s, and the corresponding filter length is 600mm. Experimental verification shows that the filter breakthrough time is 143s, with an error of 4.6%, which matches the theoretical calculation result. The time ratio between the predicted nodes is 31:37:50:32.

[0094] (5) Experiment 2 with any number of adsorption columns selected as the prediction unit

[0095] Similarly, under the above conditions, in the case of Figure 3 In the prediction system shown, valves 18 (44), 19 (45), 20 (46), 22 (48), 23 (49), 27 (53), 28 (54), 29 (55), and 30 (56) are opened. Using Equation 8, the equivalent length and breakthrough time of each prediction unit after opening the above valves are calculated, as shown in Table 9.

[0096] Table 9

[0097] 1 2 3 4 equivalent length 100 187 211 87 Theoretical breakthrough time 19 66 120 142

[0098] Prediction experiments were conducted, and the breakthrough time for each prediction unit is shown in Table 10. The predicted breakthrough time of the indicator system was 144s, with an error of less than 1% compared to the filter failure time, and it matched the theoretical calculation results. The time ratio between prediction nodes was 19:48:47:30.

[0099] Table 10

[0100] Prediction Unit 1 2 3 4 Breakthrough Time 19 67 114 144

[0101] Therefore, by changing the number and length of the adsorption columns, the remaining effective time of the filter at different time points can be predicted, thus adapting to more situations.

[0102] Example 3: This example is based on Example 1, but only the entry conditions are changed to verify the reliability of the prediction indication system under different conditions.

[0103] The schematic diagram of the device in this embodiment is as follows: Figure 2 As shown, the prediction system consists of four prediction units: the first, second, third, and fourth prediction units. Units 1-12 comprise the first to twelfth adsorption columns, units 13-27 comprise the first to fifteenth valves, unit 28 is the concentration detector, and unit 29 is the pump. The packing lengths of each adsorption column in this prediction device are shown in Table 1.

[0104] exist Figure 2 In the schematic diagram of the device shown, adsorption columns with packing lengths of 80mm, 60mm, 40mm, and 20mm are connected to conduct a predictive experiment, namely, opening the first valve 13, the fourth valve 16, the seventh valve 19, the tenth valve 22, the thirteenth valve 25, the fourteenth valve 26, and the fifteenth valve 27.

[0105] In this embodiment, NO is used as the toxic gas, with an inlet flow rate of 47.1 ml / min (half that of Example 1) and an inlet concentration of 0.4 mol / m³. 3 .

[0106] Under the above conditions, a prediction experiment was conducted on the prediction indication system. During the prediction process, the breakthrough time of each prediction unit was recorded, as shown in Table 11.

[0107] Table 11

[0108] Prediction Unit 1 2 3 4 Breakthrough Time 22 47 66 77

[0109] The predictive indication system can predict the filter failure time for a filter with a filling length of 200mm. Through experiments, the filter failure time was 72s, the predictive indication system failure time was 77s, and the prediction error was 6%. First of all, the predictive indication system can still accurately predict the filter failure time under different conditions.

[0110] Secondly, the remaining effective time of the filter under different conditions can be calculated by formula (3). After the first prediction unit breaks through, the remaining effective time is 52s by formula (3), and the actual remaining time is 55s. After the second prediction unit breaks through, the remaining time is 22s by formula (3), and the actual remaining time is 20s. Therefore, the prediction method of the present invention can predict the failure time of the filter under unknown conditions.

[0111] Through the above embodiments, the prediction indication system of the present invention has a low prediction error during the prediction process. Without considering experimental errors, the prediction method can accurately predict the filter failure time. At the same time, by adjusting the valve and connecting adsorption columns of different lengths and numbers, the remaining effective time of the filter at different time points can be predicted, thereby expanding the applicability of the prediction device and providing a basis for real-time monitoring of the filter's filtration effect and replacement of the filter with a new one.

Claims

1. A multi-site predictive filter failure time indication system, characterized in that, The indicated system includes a control system, a concentration detection system, a prediction system, and a flow regulation system. The prediction system consists of n prediction units, each consisting of m adsorption columns of equal or unequal length. All adsorption columns in all prediction units must have the same diameter. Different prediction units may have the same or different number of adsorption columns, and the lengths of the adsorption columns in different prediction units may be equal or unequal. In each prediction unit, the adsorption columns are connected in parallel, and each adsorption column is equipped with a valve. The prediction units are connected in series. Except for the last prediction unit, the main circuit at the end of each prediction unit is also connected to the inlet of the concentration detection system. Each system is equipped with a valve; the last prediction unit is directly connected to the inlet of the concentration detection system; by adjusting the opening and closing of the internal valves, adsorption columns of different lengths and numbers in each prediction unit are connected, thereby achieving the purpose of automatically switching prediction nodes and predicting the remaining time of the filter at different time points; the outlet of the concentration detection system is connected to the pump inlet, and the pump outlet is connected to the polluted environment; the control system is connected to the concentration detection system, prediction system, flow regulation system, and pump; the flow regulation system is connected to the inlet of the first prediction unit; the adsorption columns and filters are filled with the same adsorbent and work synchronously under the same operating conditions to avoid the influence of complex external conditions on the indicating system; For different applications, it is necessary to predict the remaining time of the filter at different time points. Therefore, by connecting adsorption columns of different lengths and numbers built into the indicator system, the remaining effective time of the filter at different time points can be predicted, as follows: (1) Open the corresponding valve in the prediction system according to the filter size and the predicted time node, and connect the adsorption columns of different lengths and numbers; (2) Obtain the filter operating conditions and adjust the inlet flow of the prediction system through the flow regulation system to ensure that the prediction system and the filter operating conditions are the same; the filter operating conditions refer to the inlet gas type, inlet concentration, empty bed gas velocity and operating temperature; (3) The filter and the prediction system start working simultaneously: Based on the filter operating conditions obtained in step (2), when the relationship between the packing length and the breakthrough time in the adsorption column is known for the filter under this operating condition, the remaining effective time of the filter can be calculated directly by the packing length in each prediction unit. Based on the filter operating conditions obtained in step (2), when the relationship between the packing length and breakthrough time in the adsorption column is unknown for the filter under this operating condition, then, on the basis of ensuring that the total packing length of the prediction unit is equal to the filter packing length, and according to the principle that the ratio of the breakthrough time corresponding to the same packing length to the total breakthrough time remains unchanged under different conditions, combined with the relationship between the packing length and breakthrough time in a certain adsorption column obtained in advance under experimental conditions, the remaining failure time of the filter under any condition is calculated by the following formula: （3）； Among them, t i * t represents the predicted breakthrough time of the unit under unknown conditions. b * Represents different values ​​of t under unknown conditions i * The corresponding remaining valid time; t i The value represents the predicted unit breakthrough time under experimental conditions, and T represents the filter breakthrough time under experimental conditions.

2. The multi-site predictive filter failure time indication system according to claim 1, characterized in that, n is 4 or 5, m is 2-4, and the same operating conditions for the adsorption column and filter mean that the inlet gas type, inlet concentration, empty bed gas velocity and operating temperature are the same.

3. The multi-site predictive filter failure time indication system according to claim 1, characterized in that, The flow regulation system is used to regulate the flow rate entering the indicator system, ensuring that the indicator system and the filter work synchronously. The control system is used to control the working status of each electric component and display it on the panel, as well as to store and process the predicted data. The concentration detection system has a built-in concentration sensor or an external concentration detection device.

4. The multi-site predictive filter failure time indication system according to claim 1, characterized in that, In addition to predicting filter failure time, the indicator system can also monitor the external environment in real time, including real-time monitoring of toxic gas concentration, temperature, and humidity.

5. A method for predicting the failure time of a multi-site predictive filter using the multi-site predictive filter failure time indication system according to any one of claims 1-4, characterized in that, For different applications, it is necessary to predict the remaining time of the filter at different time points. Therefore, by connecting adsorption columns of different lengths and numbers built into the indicator system, the remaining effective time of the filter at different time points can be predicted, as follows: (1) Open the corresponding valve in the prediction system according to the filter size and the predicted time node, and connect the adsorption columns of different lengths and numbers; (2) Obtain the filter operating conditions and adjust the inlet flow of the prediction system through the flow regulation system to ensure that the prediction system and the filter operating conditions are the same; the filter operating conditions refer to the inlet gas type, inlet concentration, empty bed gas velocity and operating temperature; (3) The filter and the prediction system start working simultaneously: Based on the filter operating conditions obtained in step (2), when the relationship between the packing length and the breakthrough time in the adsorption column is known for the filter under this operating condition, the remaining effective time of the filter can be calculated directly by the packing length in each prediction unit. Based on the filter operating conditions obtained in step (2), when the relationship between the packing length and breakthrough time in the adsorption column is unknown for the filter under this operating condition, then, on the basis of ensuring that the total packing length of the prediction unit is equal to the filter packing length, and according to the principle that the ratio of the breakthrough time corresponding to the same packing length to the total breakthrough time remains unchanged under different conditions, combined with the relationship between the packing length and breakthrough time in a certain adsorption column obtained in advance under experimental conditions, the remaining failure time of the filter under any condition is calculated by the following formula: （3）； Among them, t i * t represents the predicted breakthrough time of the unit under unknown conditions. b * Represents different values ​​of t under unknown conditions i * The corresponding remaining valid time; t i The value represents the predicted unit breakthrough time under experimental conditions, and T represents the filter breakthrough time under experimental conditions.

6. The method according to claim 5, characterized in that, In step (1), the filter gauge dimensions include the filter length and cross-sectional area.

7. The method according to claim 5, characterized in that, In step (1), the principle for the length and number of adsorption columns connected is to ensure that the total packing length or equivalent length of the adsorption columns is equal to the length of the filter packing.

8. The method according to claim 5, characterized in that, In step (1), the methods for selecting the length and number of adsorption columns include the following two categories: Method 1: Select each prediction unit to connect only one adsorption column, and the packing material lengths of the connected adsorption columns can be the same or different; then the relationship between the adsorption column and the prediction time is linear: （1）； Where y represents the breakthrough time and x represents the length of the adsorption column packing; Method 2: When selecting at least one prediction unit that connects two or more adsorption columns, the packing length of all connected adsorption columns within the same prediction unit must be equal. The packing lengths of adsorption columns in different prediction units may be the same or different. Multiple adsorption columns are treated as a single adsorption column for equivalent prediction. （2）； Where Z is the equivalent length of the adsorption column packing material to x, except for x; and x is the packing material length of any adsorption column within each prediction unit.

9. The method according to claim 5, characterized in that, In step (2), the inlet flow rate of the prediction system is adjusted by the flow regulation system as follows: when the diameter of the adsorption column in the indicator system is equal to the diameter of the filter, the flow rates of the indicator system and the filter are equal; if the two diameters are not equal, let the filter inlet flow rate be Q, the filter diameter be D, the adsorption column diameter be d, and the indicator system inlet flow rate be Q. y Then the flow regulation system must ensure that the flow at the inlet of the indicator system satisfies the following relationship: （4）。

Images