Carbon dioxide purification control method based on concentration prediction in cabin environment
By deploying carbon dioxide sensors and using an autoregressive integral moving average algorithm to construct a concentration prediction model in a closed cabin environment, future concentration predictions are generated. Combined with personnel activity status, control commands are generated, solving the problems of lag in carbon dioxide concentration regulation and ineffective energy consumption within the cabin, thus achieving efficient environmental control and energy management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-29
- Publication Date
- 2026-04-10
AI Technical Summary
Existing technologies suffer from problems such as delayed carbon dioxide concentration control and inefficient energy consumption in closed or semi-closed cabin environments, making it impossible to achieve a balance between energy efficiency and personnel comfort while ensuring environmental safety.
By deploying carbon dioxide concentration sensors to acquire real-time monitoring data, an autoregressive integral moving average algorithm is used to construct a concentration prediction model, generating a sequence of future concentration prediction values. Combined with the activity status of personnel inside the cabin, differentiated control commands are generated to achieve proactive control.
It effectively solves the problem of regulatory lag, improves energy utilization efficiency, reduces energy consumption during quiet periods, and achieves an intelligent balance between environmental safety and comfort.
Smart Images
Figure CN121828847A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of concentration prediction technology, specifically to a carbon dioxide purification control method based on concentration prediction in a cabin environment. Background Technology
[0002] In enclosed or semi-enclosed environments (such as spacecraft cabins, submarine compartments, high-level biosafety laboratories, and special protective vehicles), maintaining carbon dioxide concentration below a safe threshold is a fundamental requirement for ensuring personnel health and the safe operation of equipment. Current technologies generally employ a feedback control strategy based on a fixed threshold. This involves real-time monitoring of the carbon dioxide concentration within the cabin using sensors, and only activating purification equipment (such as circulating fans or regenerative carbon dioxide removal devices) when the monitored value exceeds a preset safety threshold. This passive, reactive control method inherently suffers from lag. For example, there is a significant time delay between the concentration exceeding the limit, its detection by sensors, the response of the control system, and the final purification effect produced by the actuators. During this period, the carbon dioxide concentration may have already risen to an adverse level that has a physiological impact on personnel. Furthermore, when the concentration is below the safety threshold, continuously operating equipment at a fixed power level to prevent slow accumulation leads to inefficient energy consumption. This energy inefficiency is particularly pronounced in cabin environments with limited energy supply.
[0003] On the other hand, existing technologies lack adaptability to the activity patterns of people inside the cabin. For example, during quiet periods such as when people are sleeping, the rate of carbon dioxide production is low and people's sensory requirements for airflow are different. If the same control logic is used as during active periods, energy conservation will either fail to be achieved, or noise interference may occur due to unnecessary full-power operation of equipment. Therefore, current technologies cannot achieve an intelligent balance between energy efficiency and human comfort while ensuring environmental safety. Summary of the Invention
[0004] The purpose of this invention is to provide a carbon dioxide purification control method based on concentration prediction in a cabin environment, in order to solve the problems mentioned in the background art. Specific technical problems include how to transform the lagging control based on real-time concentration thresholds into a judgment based on whether the concentration prediction value sequence exceeds the threshold, and to combine this with the personnel activity quiet period status signal for joint logical decision-making, thereby generating forward-looking and differentiated equipment control commands. This addresses the control lag problem of existing technologies relying on feedback after exceeding the standard, and the problem of ineffective energy consumption caused by the inability to identify the cabin's working quiet period.
[0005] To achieve the above objectives, the present invention provides a carbon dioxide purification control method based on concentration prediction in a cabin environment, specifically comprising the following method steps: S1. Acquire real-time monitoring data sequences from a carbon dioxide concentration sensor, wherein the process of acquiring the real-time monitoring data sequences specifically includes: A carbon dioxide concentration sensor deployed inside the cabin periodically detects the analog signal of carbon dioxide concentration at a set sampling frequency and converts the analog signal into a digital signal via an analog-to-digital converter circuit. The controller reads the discrete concentration values arranged in chronological order from the carbon dioxide concentration sensor through a communication interface and caches them in a designated storage area to form a real-time monitoring data sequence with timestamps.
[0006] Step S1, by collecting and processing carbon dioxide concentration data in the cabin in real time, provides an accurate and continuous input sequence for subsequent prediction and control, thus solving the problem that control methods need to rely on real-time monitoring data as a basis for decision-making.
[0007] S2. Input the real-time monitoring data sequence into the pre-stored concentration prediction model, and output the concentration prediction value sequence for a preset time period in the future; wherein the concentration prediction model is constructed based on the autoregressive integral moving average algorithm; the training process of the concentration prediction model specifically includes: Historical carbon dioxide concentration data of the cabin under various typical operating conditions were collected to form a training dataset. Perform stationarity testing and differencing on the training dataset; The autoregression order, difference order, and moving average order are determined by analyzing the autocorrelation plot and partial autocorrelation plot of the training dataset. The training dataset was fitted using an autoregressive integral moving average model of a given order, and the fixed coefficients of the concentration prediction model were calculated using the maximum likelihood estimation method.
[0008] The specific process of outputting the concentration prediction value sequence includes: The real-time monitoring data sequence is differentially processed; Starting with the latest actual concentration value, iterative calculations are performed using the order of the concentration prediction model and fixed coefficients. Based on the linear combination formula, the estimated concentration value at each preset time point in the future is calculated in sequence to form a sequence of concentration prediction values.
[0009] Step S2 uses a prediction model trained on historical data to extrapolate the concentration change trend over a future period based on the current monitoring sequence. This shifts the basis for control decisions from "whether the current level exceeds the standard" to "whether the level will exceed the standard in the future," thus providing a forward-looking judgment basis for solving the problem of regulatory lag.
[0010] S3. Compare the predicted concentration value sequence with the pre-stored concentration safety threshold, and select to generate a first control command sequence and a second control command sequence based on the comparison result, wherein: If it is determined that the predicted concentration value sequence will exceed the concentration safety threshold within a preset time period, a first control command sequence is generated. This first control command sequence is used to control the start of the circulating fan and to control the regeneration heater to operate at the power value calculated based on the concentration rise rate. The execution process of the first control command sequence includes: Set the status of the circulating fan to "Start"; The concentration rise rate is calculated by taking the difference between the earliest predicted value that exceeds the concentration safety threshold in the concentration prediction value sequence and the predicted value of the previous sampling point, and then dividing it by the sampling time interval. The calculated rate value is substituted into a predefined linear function to calculate the operating power value of the regenerator and generate the corresponding power control command; the predefined linear function calculates the operating power value based on the product of the base power and the proportional coefficient and the concentration rise rate.
[0011] If the predicted concentration value sequence is determined to be below the concentration safety threshold within a preset time period, and a status signal indicating a quiet period for personnel activity within the chamber is received, a second control command sequence is generated. This second control command sequence is used to control the circulating fan to operate at a preset minimum speed and to control the regeneration heater to shut down. The specific process of receiving the status signal indicating a quiet period for personnel activity includes: Monitor the silent period flag bit in the status signal register from the in-cabin schedule management system; When the time enters the predefined quiet period schedule, the quiet period flag is set to active; When the comparison result is false, a logical AND operation is performed with the value of the silent period flag. When both are true, the generation of a second control instruction sequence is triggered. The execution process of the second control instruction sequence includes: Set the target speed of the circulating fan to the preset minimum speed value; Set the power target value of the regenerative heater to zero to shut down the regenerative heater.
[0012] Step S3 involves making a combined logical judgment based on the concentration prediction results and the activity status inside the chamber, and intelligently generating differentiated control commands: when the concentration exceeds the standard, the equipment is started in advance and the power is adjusted according to the predicted rate of increase in concentration to cope with the upcoming high load; when the concentration is predicted to be safe and in a quiet period, a low-energy consumption mode is executed, directly addressing the two specific technical problems of control lag and ineffective energy consumption during the quiet period.
[0013] S4. Send the generated first control instruction sequence or second control instruction sequence to the corresponding execution unit, specifically including: The controller sends the generated first control command sequence or second control command sequence in a defined binary encoding format to the corresponding execution unit through its communication interface; wherein, the first control command sequence is sent to the circulating fan and the regenerative heater to control the fan to start and operate at the calculated power; the second control command sequence is sent to the circulating fan and the regenerative heater to control the fan to maintain the minimum speed and shut down the heater.
[0014] Step S4 involves accurately and promptly sending the generated forward-looking and differentiated control commands to the actuators such as the circulating fan and regenerative heater, ensuring that the optimized control strategy based on prediction and logical decision-making is physically realized, ultimately achieving the goal of improving control timeliness and energy-saving effect.
[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention transforms lagging control based on real-time concentration into forward control based on a predicted future concentration sequence, and combines this with joint decision-making based on the state of the chamber during the quiet period. This enables intelligent pre-start and on-demand power adjustment of the purification equipment. This effectively overcomes the control delay problem caused by the reliance on feedback after exceeding the standard in traditional methods. At the same time, it automatically switches to a low-energy operation mode when it is confirmed that there is no one in the chamber and the concentration is safe, fundamentally solving the problem of ineffective energy consumption during the quiet period. Thus, it significantly improves energy utilization efficiency while ensuring environmental indicators. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the overall method steps of the present invention; Figure 2 This is the core flowchart of step S3 of the present invention. Detailed Implementation
[0017] The technical solutions in 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 embodiments of the present invention, and not all embodiments. 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.
[0018] Next, please refer to Figure 1 The purpose of this embodiment is to provide a carbon dioxide purification and control method based on concentration prediction in a cabin environment, which includes the following steps: S1. A carbon dioxide concentration sensor deployed at a specific location inside the cabin periodically detects the analog signal of the carbon dioxide concentration in the environment at a set sampling frequency, and converts the analog signal into a digital signal via an analog-to-digital converter circuit. Subsequently, the controller continuously reads these discrete concentration values arranged in chronological order from the sensor through its communication interface (such as CAN bus, RS-485 or Ethernet), and caches them in a designated storage area to form an ordered, timestamped real-time monitoring data sequence, thereby providing continuous and dynamic raw data input for subsequent concentration prediction.
[0019] S2. Input the real-time monitoring data sequence into the pre-stored concentration prediction model, and output the concentration prediction value sequence for a preset time period in the future. The pre-stored concentration prediction model is built based on the autoregressive integral moving average algorithm, and its offline training process is as follows: First, historical carbon dioxide concentration data of the cabin under various typical operating conditions are collected to form a training dataset. Then, the stationarity of the training dataset is tested and differencing is performed to ensure that it meets the application conditions of the concentration prediction model. Next, the core parameters of the model, namely the autoregressive order, the differencing order, and the moving average order, are determined by analyzing the autocorrelation plot and partial autocorrelation plot of the training dataset. Finally, the training dataset is fitted with an autoregressive integral moving average model of determined order, and the fixed coefficients of the concentration prediction model are calculated by the maximum likelihood estimation method, thereby completing the construction of the concentration prediction model and pre-storing it in the controller.
[0020] The concentration prediction model outputs a sequence of predicted concentration values for a preset future time period based on real-time monitoring data. It is a deterministic online rolling prediction process, as detailed below: The concentration prediction model takes the received real-time monitoring data sequence as input and first performs the same differential processing as in the training phase. Then, using the order of the concentration prediction model determined in the training phase and the calculated fixed coefficients, it performs iterative calculations starting from the latest actual concentration value. This calculation is based on the inherent linear combination formula of the concentration prediction model, weighting and summing the residuals of observed values and predicted values at several past time points to sequentially calculate the concentration prediction value at each preset time point in the future. These time-series generated prediction values together constitute the concentration prediction value sequence for a preset time period in the future. The preset time period is a key system parameter, and its length is mainly determined by the total response delay of the system, specifically including: This time period should be greater than or equal to the total time required from the generation of control commands by the controller to the full activation of the actuators (circulating fan, regenerating heater) and their beginning to significantly affect the carbon dioxide concentration inside the chamber. This includes the communication transmission time of the control commands, the mechanical response and power ramp-up time of the actuators, and the physical process time of air circulation and mixing. By setting it in this way, it is ensured that the forward-looking control decisions made based on the predicted concentration value at the end of this time period can just in time offset the rising trend of the concentration, thereby enabling the purification system to take effect before the concentration exceeds the standard, achieving a fundamental shift from delayed control to early intervention. The specific value of this time period can be experimentally determined and configured according to the volume of different chambers, air circulation characteristics, and equipment performance. It is usually a settable time value, such as 5 minutes, 10 minutes, or 30 minutes.
[0021] S3, please refer to Figure 2 The concentration prediction sequence is compared with pre-stored concentration safety thresholds, specifically including: The controller sequentially scans each predicted value in the concentration prediction sequence and compares it with a pre-stored single concentration safety threshold. The core of this comparison is determining whether any predicted value in the sequence exceeds the safety threshold. The result of this comparison is a Boolean value. If true, it triggers the path to generate the first control command sequence; if false, it further performs a logical AND operation with a state signal indicating a quiet period for personnel activity within the chamber. Only when both conditions are true is it triggers the path to generate the second control command sequence. This is a deterministic conditional branch logic, entirely driven by the comparison result between the predicted data and a fixed threshold. The specific process of performing a logical AND operation with the state signal indicating a quiet period for personnel activity within the chamber is as follows: A status signal register from the cabin schedule management system or personnel sensing system is continuously monitored. A specific binary bit in this register is defined as the silent period flag bit. When the time enters the predefined silent period schedule (such as the bedtime period) or the personnel sensor detects that there is no activity in the cabin for more than a preset time, this flag bit is set to 1 by hardware or upper-level software; otherwise, it is 0. Subsequently, during the decision cycle for generating control instructions, the controller's arithmetic logic unit performs a deterministic bitwise AND operation. It multiplies the judgment result that the aforementioned concentration prediction value sequence has not exceeded the threshold (which has been converted to the Boolean value True, with an integer value of 1) with the value of the silent period flag bit (1 or 0) read from the status signal register (i.e., the physical implementation of logical AND). The condition of the entire AND judgment is satisfied only if and only if the result of the multiplication is 1 (i.e., both input conditions are true), thereby triggering the path to generate the second control instruction sequence. This is a deterministic judgment guaranteed by hardware instructions, an either-or decision.
[0022] When the comparison result is true, the first control instruction sequence is generated, which specifically includes: First, an instruction is immediately generated to set the circulating fan to start. Second, the concentration rise rate is calculated by taking the difference between the earliest predicted value exceeding the safety threshold in the concentration prediction value sequence and the predicted value of the previous sampling point, and then dividing by the sampling time interval. Then, the calculated rate value is substituted into a predefined linear function P=P_base+k*rate (where P_base is the base power and k is the proportional coefficient) to calculate the operating power value P of the regenerative heater. Finally, the corresponding power control instruction is generated, which is arranged in the execution sequence to form the first control instruction sequence.
[0023] When the comparison result is false and the status signal of the personnel activity silence period is true, a second control command sequence is generated, specifically including: The second control command sequence contains only two control commands. The first command sets the target speed of the circulating fan to a preset minimum speed value, which is used to control the circulating fan to operate at the preset minimum speed. The minimum speed value is only used to maintain basic air circulation in the cabin. The second command sets the target power value of the regenerative heater directly to zero, that is, issues a shutdown command. These two commands are encapsulated into a command sequence, which aims to switch the system to the maintenance operation mode with the lowest energy consumption.
[0024] S4. Send the generated first control instruction sequence or second control instruction sequence to the corresponding execution unit, specifically including: The controller sends the first or second control command sequence generated in step S3 to the corresponding execution components, including the circulating fan and the regenerative heater, in an unambiguous binary encoding format via its communication interface (such as CAN bus, RS-485, or Ethernet). After the first control command sequence is sent, the circulating fan will start according to the command, and the regenerative heater will operate according to the calculated power value P. After the second control command sequence is sent, the circulating fan will switch to the preset minimum speed, and the regenerative heater will receive the shutdown command and stop working, thereby achieving precise and real-time control of the entire environmental control system.
[0025] It should be noted that the controller in this embodiment, as the core processing unit, is responsible for periodically reading and caching the real-time monitoring data sequence from the carbon dioxide concentration sensor through its communication interface, calling the pre-stored concentration prediction model to generate a concentration prediction value sequence for future time periods, comparing the prediction sequence with the pre-stored safety threshold, and making logical judgments by comprehensively considering the status signals of personnel activity quiet periods. Finally, it generates the first or second control command sequence based on the decision result and sends the corresponding control commands to the corresponding execution components of the circulating fan and regeneration heater through the communication interface.
[0026] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.
Claims
1. A carbon dioxide purification control method based on concentration prediction in a cabin environment, characterized in that, The methods and steps include the following: S1. Acquire real-time monitoring data sequences from the carbon dioxide concentration sensor; S2. Input the real-time monitoring data sequence into the pre-stored concentration prediction model and output the concentration prediction value sequence for a preset time period in the future. S3. Compare the predicted concentration value sequence with the pre-stored concentration safety threshold, and select to generate a first control command sequence and a second control command sequence based on the comparison result, wherein: If it is determined that the concentration prediction value sequence will exceed the concentration safety threshold within the preset time period, a first control command sequence is generated. The first control command sequence is used to control the start of the circulating fan and control the regeneration heater to operate at the power value calculated according to the concentration rise rate. If it is determined that the concentration prediction value sequence is lower than the concentration safety threshold within the preset time period, and a status signal indicating that the cabin is in a quiet period of personnel activity is received, then a second control command sequence is generated. The second control command sequence is used to control the circulating fan to run at a preset minimum speed and control the regeneration heater to shut down. S4. Send the generated first control instruction sequence or second control instruction sequence to the corresponding execution unit.
2. The carbon dioxide purification control method based on concentration prediction in a cabin environment according to claim 1, characterized in that, The process of acquiring the real-time monitoring data sequence specifically includes: A carbon dioxide concentration sensor deployed inside the cabin periodically detects analog signals of carbon dioxide concentration values at a set sampling frequency, and converts the analog signals into digital signals via an analog-to-digital converter circuit. The controller reads discrete concentration values arranged in chronological order from the carbon dioxide concentration sensor through a communication interface and caches them in a designated storage area to form a real-time monitoring data sequence with timestamps.
3. The carbon dioxide purification control method based on concentration prediction in a cabin environment according to claim 1, characterized in that, The concentration prediction model is constructed based on the autoregressive integral moving average algorithm.
4. The carbon dioxide purification control method based on concentration prediction in a cabin environment according to claim 3, characterized in that, The training process of the concentration prediction model specifically includes: Historical carbon dioxide concentration data of the cabin under various typical operating conditions were collected to form a training dataset. Perform stationarity testing and differencing on the training dataset; The autoregression order, difference order, and moving average order are determined by analyzing the autocorrelation plot and partial autocorrelation plot of the training dataset. The training dataset was fitted using an autoregressive integral moving average model of a given order, and the fixed coefficients of the concentration prediction model were calculated using the maximum likelihood estimation method.
5. The carbon dioxide purification control method based on concentration prediction in a cabin environment according to claim 1, characterized in that, The specific process of outputting the concentration prediction value sequence includes: The real-time monitoring data sequence is differentially processed; Starting with the latest actual concentration value, iterative calculations are performed using the order of the concentration prediction model and fixed coefficients. Based on the linear combination formula, the estimated concentration value at each preset time point in the future is calculated in sequence to form a sequence of concentration prediction values.
6. The carbon dioxide purification control method based on concentration prediction in a cabin environment according to claim 1, characterized in that, The process of receiving the status signal during the quiet period of personnel activity specifically includes: Monitor the silent period flag bit in the status signal register from the in-cabin schedule management system; When the time enters the predefined quiet period schedule, the quiet period flag is set to active; When the comparison result is false, a logical AND operation is performed with the value of the silent period flag bit. When both are true, the generation of the second control instruction sequence is triggered.
7. The carbon dioxide purification control method based on concentration prediction in a cabin environment according to claim 1, characterized in that, The execution process of the first control instruction sequence includes: Set the status of the circulating fan to "Start"; The concentration rise rate is calculated by taking the difference between the earliest predicted value exceeding the concentration safety threshold in the concentration prediction value sequence and the predicted value of the previous sampling point, and then dividing it by the sampling time interval. The calculated rate value is substituted into a predefined linear function to calculate the operating power value of the regenerative heater, and the corresponding power control command is generated.
8. The carbon dioxide purification control method based on concentration prediction in a cabin environment according to claim 7, characterized in that, The predefined linear function calculates the operating power value based on the product of the base power and the proportional coefficient with the concentration rise rate.
9. The carbon dioxide purification control method based on concentration prediction in a cabin environment according to claim 1, characterized in that, The execution process of the second control instruction sequence includes: Set the target speed of the circulating fan to the preset minimum speed value; Set the power target value of the regenerative heater to zero to shut down the regenerative heater.
10. The carbon dioxide purification control method based on concentration prediction in a cabin environment according to claim 1, characterized in that, The process of sending the generated control command sequence to the corresponding execution unit in step S4 specifically includes: The controller sends the generated first control command sequence or second control command sequence in a defined binary encoding format to the corresponding execution unit through its communication interface; wherein, the first control command sequence is sent to the circulating fan and the regenerative heater to control the fan to start and operate at the calculated power; the second control command sequence is sent to the circulating fan and the regenerative heater to control the fan to maintain the minimum speed and shut down the heater.