Cascade control system for controlling humidity level and dehumidification system
The MPC module with machine learning enhances dehumidifier control by predicting humidity levels and adjusting temperature setpoints, improving accuracy and stability while reducing computational burden and safety risks.
Patent Information
- Application Number
- JP2025108987
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-06-28
- Filing Date
- 2025-06-27
- Publication Date
- 2026-01-19
AI Technical Summary
Tuning PID parameters for dehumidifier control systems is time-consuming and highly subjective, leading to suboptimal performance due to the complexity of humidity control influenced by multiple environmental factors.
Implementing a model predictive control (MPC) module with machine learning algorithms to predict future humidity levels and adjust temperature setpoints, integrating sensor data to capture interdependencies and simplify the control system.
The MPC module provides more accurate, stable, and faster humidity control, eliminating the need for dedicated safety mechanisms and optimizing energy efficiency.
Smart Images
Figure 2026008976000001_ABST
Abstract
Description
[Technical Field]
[0001] The present disclosure relates to a control system for a dehumidifier, and more particularly to an improved cascade control system with a model predictive control module. [Background technology]
[0002] Fine-tuning a dehumidifier's control system to achieve optimal performance, such as maintaining a desired setpoint, can be difficult due to the many factors that affect humidity, such as the temperature and airflow characteristics of the local environment. The control system must respond quickly enough to changes while also limiting the amount of unwanted humidity fluctuations in an attempt to maintain the setpoint. Control systems typically include a series of proportional-integral-derivative (PID) controllers that determine the behavior of the control system. Summary of the Invention [Problem to be solved by the invention]
[0003] Tuning PID parameters is time-consuming and highly subjective, with each expert having their own unique approach, which can result in suboptimal parameters and settings. Therefore, there is a need in the art for an improved control system for dehumidifiers. [Means for solving the problem]
[0004] The gist of the present disclosure is to improve a cascade PID controller for a humidifier by introducing a model predictive control (MPC) module that can leverage its predictions to reach the setpoint more quickly with less fluctuation around the setpoint. By introducing artificial intelligence into the MPC module, for example, utilizing machine learning (ML) algorithms such as neural networks for prediction and / or setpoint generation, the control system can be tuned to capture the complexity of all the interdependencies between the parameters that contribute to the prediction and / or setpoint generation.
[0005] Specifically, the present disclosure relates to a cascade control system for a dehumidifier for controlling humidity levels. The dehumidifier includes a desiccant rotor. The desiccant rotor includes a desiccant material. The dehumidifier is configured to cause a process air stream and a reactivated air stream to traverse separate axially extending passages through the desiccant material. Each air stream has an inlet air flow as it enters the desiccant material and an outlet air flow as it exits the desiccant material. The inlet reactivated air is configured to remove moisture from the desiccant material when heated. The cascade control system includes a heater for adjusting the humidity level of the outlet process air by heating the inlet reactivated air. The cascade control system further includes a primary control module for controlling the humidity level by controlling at least one operating parameter of the dehumidifier. The cascade control system also includes a secondary control module for controlling the temperature of the inlet reactivated air to match a desired temperature setpoint by controlling at least one operating parameter of the heater. The primary control module is configured to provide as an output a desired temperature setpoint for the secondary control module while adjusting the humidity level to match the desired humidity setpoint. The primary control module is a model predictive control (MPC) module configured to predict future humidity levels based on historical humidity level data and the historical desired temperature setpoint data provided to the secondary control module, and to calculate the desired temperature setpoint for the secondary control module based on the predicted humidity level.
[0006] This provides a cascade control system that provides more accurate, more stable, and faster control of humidity levels compared to a corresponding cascade PID control system. A further potential technical effect and advantage is that the cascade control system configuration may be simplified relative to the prior art. Specifically, many prior art cascade control systems require a dedicated safety mechanism, such as a switch, to prevent the heater from overheating. The disclosed cascade control system does not require such a dedicated safety mechanism.
[0007] According to one embodiment, the MPC module is configured to utilize machine learning (ML) algorithms to predict future humidity levels.
[0008] The use of machine learning means that predictive models do not have to be explicitly modeled, but can be discovered by providing the machine learning algorithm with appropriate training data. A further technical effect is that the machine learning algorithm enables the MPC module, and therefore the cascaded control system, to capture and implicitly model complex interdependencies, e.g., integrating sensor data from multiple sensors.
[0009] According to one aspect, the secondary control module includes at least one PID controller.
[0010] According to certain embodiments, the MPC module is further configured to predict future humidity levels and / or calculate a desired temperature setpoint for the secondary control module based on sensor data related to parameters affecting the humidity level being controlled. By integrating additional sensor data, such as the humidity level of the incoming process air, the model predictions of the MPC module provide improved control performance, such as more responsive control.
[0011] According to one embodiment, the MPC module includes a prediction module and an optimization module. The prediction module is configured to predict future humidity levels. Predicting future humidity levels further includes generating a plurality of consecutive future humidity levels and generating a plurality of future desired temperature setpoints. The optimization module is configured to calculate a desired temperature setpoint for the secondary control module. The calculation of the desired temperature setpoint is based on at least one of a humidity level control performance metric using the plurality of consecutive future humidity levels, a desired temperature setpoint stability metric using the plurality of future desired temperature setpoints, and an energy efficiency performance metric using the plurality of future desired temperature setpoints.
[0012] This allows a compromise to be reached between three (potentially conflicting) objectives: achieving a humidity level close to the desired humidity setpoint, stabilizing the heater temperature, and achieving optimal energy efficiency.
[0013] According to one embodiment, prediction of future humidity levels is performed using a linear model, which makes the optimization problem a convex optimization problem.
[0014] According to certain aspects, calculating the desired temperature setpoint is determined from a closed-form solution, which significantly accelerates optimization and reduces computational burden.
[0015] According to one aspect, the cascade control system further includes a primary PID control module configured to control the humidity level by controlling at least one operating parameter of the dehumidifier and to provide a desired temperature setpoint for the secondary control module when a predetermined criterion is met.
[0016] By providing a cascade control system with a PID control module that mimics the MPC module, the primary PID control module can serve as a reference and provide training data for the machine learning algorithm of the MPC module. When the MPC module is trained to a level where the MPC module and secondary control module function similarly to the primary PID control module, the primary PID control module can be switched off. Similarly, if there is a sudden change in the operating environment that causes the MPC module and secondary control module to function poorly, the cascade control system can switch to using the primary PID control module to collect training data for the new operating environment. The primary PID control module thereby serves as both a fallback and a mechanism for any machine learning algorithms to obtain relevant training data.
[0017] According to one embodiment, the cascade control system is further configured to adjust the desired temperature setpoint of the secondary control module based on the temperature and / or humidity level of the effluent reactivation air.
[0018] According to one aspect, the cascade control system further includes a tertiary control module for controlling the temperature of the effluent reactivation air by adjusting a desired temperature setpoint of the secondary control module based on a difference between the temperature of the effluent reactivation air and a predetermined threshold temperature.
[0019] The reason for this is safety and efficiency. The idea is that as long as the temperature of the effluent reactivation air is below a predefined threshold, then the tertiary control module is inactive. Once the temperature exceeds the threshold, then the tertiary control module is activated and adjusts (lowers) the desired temperature setpoint of the secondary control module.
[0020] According to one aspect, the cascade control system further includes a quaternary control module configured to receive a desired room humidity setpoint associated with the humidity level of a room to which the effluent process air is supplied, receive a current humidity level for the room, and generate a desired humidity setpoint for the primary control module.
[0021] The cascade control system thereby has three nested loops - an outermost loop for controlling the humidity level of the room, an inner loop for controlling the humidity level of the effluent treatment air, and an innermost loop for controlling the temperature of the effluent reactivation air.
[0022] The advantage of doing this instead of directly controlling the room humidity is that the room humidity control device that now controls the dry air humidity is simplified, and the relationship between the humidity of the outgoing process air and the room humidity is simpler than the relationship between the dehumidifier actuator and the room humidity.
[0023] Another advantage is that the control loop for the outflow process air only involves the dehumidifier and not the room, allowing the dehumidifier itself to operate in an energy-efficient manner while the external (room humidity) control device can be configured to achieve the goal of ensuring the room has the desired humidity.
[0024] The present disclosure further relates to a dehumidification system for controlling humidity levels, comprising a dehumidifier and a cascade control system for controlling humidity levels as described above and below, the dehumidification system having all of the technical effects and advantages of the cascade control system. [Brief explanation of the drawings]
[0025] [Figure 1] Diagram showing an example of a dehumidifier that can use a cascade control system [Figure 2] FIG. 1 illustrates the disclosed cascade control system. [Figure 3] Diagram showing a dehumidification system DETAILED DESCRIPTION OF THE INVENTION
[0026] 1 shows an example of a dehumidifier 102 that can use the disclosed cascade control system. The dehumidifier 102 includes a desiccant rotor 103. The desiccant rotor contains a desiccant material. In the illustrated example, the dehumidifier includes a drive motor 109 arranged to rotate the desiccant rotor 103.
[0027] The dehumidifier 102 is configured to direct the treatment air streams 104a-c and the reactivated air streams 105a-d through separate axially extending passages 106a, 106b through the desiccant material. Each air stream has an inlet air flow 104a, 105b as it enters the desiccant material and an outlet air flow 104b, 105c as it exits the desiccant material. The air streams are preferably physically separated, for example, by passage guides 107a, 107b.
[0028] In practice, the dehumidification process begins by passing inlet process air 104a through a desiccant material, thereby causing the desiccant material to absorb and / or adsorb moisture. For simplicity, the term adsorption will be used throughout the description to describe the capture of moisture by the desiccant material, but it should be understood that absorption can be a complementary or even primary process by which moisture is captured. The process air then exits the desiccant material as dehumidified process air, referred to herein as outlet process air 104b or dry air. The outlet process air 104b can then be directed 104c to its intended destination, such as a room, by, for example, a fan 108b.
[0029] The amount of moisture already adsorbed by the desiccant material influences the amount of moisture adsorbed in the incoming process air. Therefore, a mechanism exists for removing moisture from the desiccant material using a reactivation air stream. Specifically, hot air is directed at the desiccant material, where the heat causes the adsorbed moisture to desorb from the desiccant material and be carried away by the reactivation air stream. In other words, the heater is positioned to adjust the humidity level of the outgoing process air 104b by heating the incoming reactivation air 105b. When the adsorbed moisture is removed from the desiccant material by the reactivation air stream, the desiccant material can more effectively dehumidify the incoming process air, thereby indirectly adjusting the humidity level.
[0030] The reactivated air can be taken from air diverted from the process air stream or from air recirculated from a space where the outflowing process air is used to control its humidity level. The reactivated air 105a is heated in a heater 110 so that the inflowing reactivated air 105b is heated enough that moisture is released from the desiccant material and carried away in the outflowing reactivated air 105c. The outflowing reactivated air 105c, also referred to herein as humid air, can then be removed 105d using, for example, a fan 108a.
[0031] 2 illustrates a disclosed cascade control system 200 of a dehumidifier 202 for controlling humidity levels. According to one embodiment, the controlled humidity level is the humidity level of the outgoing process air 204b. According to one embodiment, the controlled humidity level is the humidity level of a room, such as room 204c, to which the outgoing process air is directed.
[0032] The dehumidifier 202 includes a desiccant rotor. The desiccant rotor contains a desiccant material. The dehumidifier is configured to direct process air streams 204a-204c and reactivated air streams 205a-b across separate axially extending passages through the desiccant material, with each air stream having an inlet air stream 204a, 205a as it enters the desiccant material and an outlet air stream 204b, 205b as it exits the desiccant material, respectively. The inlet reactivated air stream 205a is configured to remove moisture from the desiccant material when heated. The cascade control system 200 includes a heater 210 for adjusting the humidity level of the outlet process air by heating the inlet reactivated air stream.
[0033] Cascade control system 200 further includes a primary control module 220 for controlling the humidity level by controlling at least one operating parameter of dehumidifier 202. Cascade control system 200 also includes a secondary control module 230 for controlling the temperature of inlet reactivated air 205a to match a desired temperature setpoint SP2 by controlling at least one operating parameter of heater 210. According to one embodiment, secondary control module 230 includes at least one PID controller 232.
[0034] The primary control module 220 is configured to provide as an output the desired temperature setpoint SP2 of the secondary control module while adjusting the humidity level to match the desired humidity setpoint SP1.
[0035] The primary control module 220 is a model predictive control (MPC) module configured to predict future humidity levels based on historical humidity level data and historical desired temperature setpoint data provided to the secondary control module, and to calculate a desired temperature setpoint SP2 for the secondary control module 230 based on the predicted humidity level.
[0036] According to one embodiment, the MPC module is further configured to predict future humidity levels and / or calculate a desired temperature setpoint SP2 for the secondary control module based on sensor data related to parameters affecting the humidity level being controlled, one example of which is the humidity level of the incoming process air.
[0037] According to one embodiment, MPC module 220 is configured to utilize machine learning (ML) algorithms to predict future humidity levels. In one example, the machine learning algorithm is a single ML algorithm configured to specify a desired humidity setpoint SP1 (with at least one measurement related to the humidity level to be adjusted as input, and optionally sensor data related to parameters affecting the humidity level being controlled) and to calculate a desired temperature setpoint SP2 as output. In some other examples, MPC module 220 includes multiple modules, each potentially implementing a respective ML algorithm, such as prediction module 222 and optimization module 224 described above and below.
[0038] According to one embodiment, the MPC module includes a prediction module 222 and an optimization module 224. The prediction module 222 is configured to predict future humidity levels. Predicting future humidity levels further includes generating a plurality of consecutive future humidity levels and generating a plurality of future desired temperature setpoints. The optimization module 224 is configured to calculate a desired temperature setpoint for the secondary control module. The calculation of the desired temperature setpoint is based on at least one of a humidity level control performance metric using the plurality of consecutive future humidity levels, a desired temperature setpoint stability metric using the plurality of future desired temperature setpoints, and an energy efficiency performance metric using the plurality of future desired temperature setpoints.
[0039] According to one embodiment, the predictor is a machine learning model that predicts humidity levels up to k steps ahead using historical data of humidity levels and desired temperature setpoints.
[0040] Once a model or predictor is found, an MPC control strategy can be used. Specifically, the optimization problem
[0041]
number
[0042] is solved at each time step t, and then the cascade control system 200 applies the initially calculated control action u t The same procedure is repeated for subsequent time steps, where TIFF2026008976000003.tif6114
[0043] is the humidity level forecast for n sample times ahead, and u t+n is the desired temperature setpoint n sample times ahead, and y SP is the desired humidity level setpoint and F is TIFF2026008976000004.tif6114
[0044] is a function that predicts u min and u max are the minimum and maximum levels of the desired temperature setpoint, and α1, α2, and α3 are arbitrary parameters between 0 and 1.
[0045] Solving the optimization problem means finding control actions to achieve three potentially conflicting objectives. The first objective is to find the optimum humidity level prediction and the desired setpoint y. SP The second objective is to keep the desired temperature setpoint command u stable, which is achieved by imposing a penalty on its changes. The third objective is to optimize energy efficiency. Adjusting α1, α2, and α3 allows for a compromise among these goals. Note that any of the parameters α1, α2, and α3 can be set to 0 to optimize only one or two of the objectives.
[0046] Therefore, according to one embodiment, the optimization module 224 is configured to calculate a desired temperature setpoint for the secondary control module. The calculation of the desired temperature setpoint involves simultaneously minimizing the cumulative difference between each future humidity level and the desired humidity setpoint, the cumulative difference between successive pairs of future desired temperature setpoints, and the cumulative sum of the squared or absolute future desired temperature setpoints. If responsiveness is prioritized without regard to energy efficiency, the optional parameter α3, which determines the relative importance of energy efficiency, can be set to 0. Therefore, in one example, the calculation of the desired temperature setpoint involves simultaneously minimizing the cumulative difference between each future humidity level and the desired humidity setpoint, and the cumulative difference between successive pairs of future desired temperature setpoints.
[0047] The data required to train the model can be obtained by applying a random signal to a desired temperature setpoint and collecting the resulting humidity level, such as the humidity level of the outgoing process air or the humidity level of the room into which the outgoing process air is directed. Instead of an MPC module, the model can also be adapted using data collected when running the system with a standard PID controller. According to certain embodiments, predictions of future humidity levels are made using a linear model. Linear models are often very effective in practice and transform the optimization problem for controlling humidity levels into a convex optimization problem. In some instances, the use of a linear model makes it possible to find a closed-form solution to the optimization problem. Therefore, according to certain embodiments, the calculation of the desired temperature setpoint is determined from the closed-form solution. The closed-form solution significantly reduces both the time it takes to solve the optimization problem and the computational resources required to do so, resulting in reduced energy consumption.
[0048] If the MPC module is configured to utilize an ML algorithm, the cascade control system can include a PID control module configured to perform the tasks of the MPC module. This additional PID control module can then serve as a reference for the MPC module, providing training data for the ML algorithm to improve the performance of the MPC module to a predetermined minimum performance level before the MPC module takes over the PID control module's role. Thus, according to one embodiment, the cascade control system further includes a primary PID control module 240 configured to control the humidity level by controlling at least one operating parameter of the dehumidifier and to provide a desired temperature setpoint SP2 for the secondary control module when a predetermined criterion is met. As noted above, in one example, the predetermined criterion relates to the relative control performance of the PMC module compared to the control performance of the primary PID control module 240.
[0049] According to one embodiment, the predetermined criteria relate to absolute control performance criteria. For example, if there is a sudden change in the operating environment that causes the MPC module and secondary control module to perform poorly, the cascade control system can switch to using the primary PID control module to collect training data for the new operating environment. The primary PID control module thereby serves as both a fallback and a mechanism for any machine learning algorithms to obtain relevant training data.
[0050] According to one embodiment, the cascade control system is further configured to adjust the desired temperature setpoint PS2 of the secondary control module based on the temperature and / or humidity level of the outflow reactivated air 205b. According to a further embodiment, the system is configured to provide the temperature and / or humidity level of the outflow reactivated air 205b as an input to the MPC module 220. In one example, the temperature and / or humidity level of the outflow reactivated air 205b is input to a prediction module 222 of the MPC module 220, which is further configured to predict future humidity levels based on the temperature and / or humidity level of the outflow reactivated air 205b.
[0051] Alternatively, the system includes a dedicated controller configured to adjust the desired temperature setpoint SP2 of the secondary control module based on the temperature and / or humidity level of the effluent reactivation air. Therefore, according to one embodiment, the cascade control system further includes a tertiary control module 250 for controlling the temperature of the effluent reactivation air by adjusting the desired temperature setpoint SP2 of the secondary control module based on the difference between the temperature of the effluent reactivation air and a predetermined threshold temperature. This provides a safety mechanism to prevent overheating of the system, which would result in the effluent reactivation air having an undesirable temperature, and allows the system to continue operating uninterrupted, thereby improving efficiency. In one example, the setpoint for the control of the effluent reactivation air is the difference between the temperature of the effluent reactivation air and the predetermined threshold temperature. According to one embodiment, the predetermined threshold temperature is selected from the range of 40°C to 80°C. According to a further embodiment, the predetermined threshold temperature is selected from the range of 50°C to 70°C. In one example, the predetermined threshold temperature is set at 60°C.
[0052] In one example, it is desirable to control the humidity level of a room to which the outflow process air is supplied. Such a room may occasionally be connected to a potential airflow that imposes a humidity load 211 on the room, such as when a door is opened and the room's air is exposed to air from an adjacent room or hallway. The desired room humidity level can be controlled by nesting three control loops in a cascade control system. The outermost loop controls the humidity level of the room, the middle loop controls the humidity level of the outflow process air, and the innermost loop controls the temperature of the outflow reactivation air. Thus, according to one embodiment, the cascade control system further includes a quaternary control module 260. The quaternary control module 260 is configured to receive a desired room humidity setpoint SP3 related to the humidity level of the room 270 to which the outflow process air is supplied, receive the current humidity level of the room, and generate a desired humidity setpoint SP1 for the primary control module. In one example, the quaternary control module is a model predictive control (MPC) module.
[0053] FIG. 3 illustrates a dehumidification system 3000 for controlling humidity levels. The dehumidification system includes a dehumidifier 302 and a cascade control system for controlling humidity levels, as described above and below. Reference numerals in FIG. 3 have the same meanings as corresponding reference numerals in FIG. 2. The dehumidifier 302 is configured to operate according to the principles described with respect to FIG. 1. Specifically, the dehumidifier 302 includes a desiccant rotor. The desiccant rotor includes a desiccant material. The dehumidifier is configured to cause process air streams 304a-c and reactivated air streams 305a-b to traverse separate axially extending passages through the desiccant material, each air stream having an inlet air stream 304a, 305a upon entering the desiccant material and an outlet air stream 304b, 305b upon exiting the desiccant material, respectively. The inlet reactivated air stream 305a is configured to remove moisture from the desiccant material when heated. Since the dehumidification system is equipped with a cascade control system as described above and below, the dehumidification system has the same technical effects and advantages. [Explanation of symbols]
[0054] 102, 202, 302 Dehumidifier 103 Desiccant Rotor 104a-c, 204a-c, 304a-c Processed air flow 105a-d, 205a-b, 305a-b Reactivation air flow 106a, 106b aisle 107a, 107b aisle guide 108a, 108b Fan 200, 300 Cascade Control System 210, 310 heater 220, 320 Primary Control Module, MPC Module 222, 322 Prediction module 224, 324 Optimization Module 230, 330 Secondary Control Module 232, 332 PDI control device 250, 350 Tertiary Control Module 260, 360 Fourth Control Module 3000 Dehumidification System
Claims
1. A cascade control system (200, 300) for a dehumidifier (102, 202, 302) for controlling humidity levels, the dehumidifier (102, 202, 302) comprising a desiccant rotor (103), the desiccant rotor (103) containing a desiccant material, the dehumidifier configured to cause a process air stream (104a-c) and a reactivated air stream (105a-d) to traverse separate axially extending passages (106a, 106b) through the desiccant material, each air stream having an inlet air (104a, 105b) upon entering the desiccant material and an outlet air (104b, 105c) upon exiting the desiccant material, the inlet reactivated air (105b) configured to remove moisture from the desiccant material when heated, the cascade control system comprising: a heater (110, 210, 310) for adjusting the humidity level of the outgoing treatment air (104b) by heating the incoming reactivation air (105b); a primary control module (220, 320) for controlling said humidity level by controlling at least one operating parameter of said dehumidifier (102, 202, 302); a secondary control module (230, 330) for controlling the temperature of said incoming reactivation air (105b) to match a desired temperature setpoint (SP2) by controlling at least one operating parameter of said heater (110, 210, 310); It is equipped with the primary control module (220, 320) is configured to provide as an output a desired temperature setpoint (SP2) of the secondary control module while adjusting the humidity level to match a desired humidity setpoint (SP1); The primary control module (220, 320) is a model predictive control (MPC) module configured to predict future humidity levels based on historical data of the humidity levels and historical data of the desired temperature setpoints provided to the secondary control module, and to calculate the desired temperature setpoints (SP2) of the secondary control modules (230, 330) based on the predicted humidity levels.
2. The cascade control system of claim 1 , wherein the MPC module (220, 320) is configured to utilize a machine learning (ML) algorithm to predict the future humidity levels.
3. The cascade control system of claim 1 or 2, wherein the secondary control module (230, 330) includes at least one PID controller (232, 332).
4. 4. The cascade control system of claim 1, wherein the MPC module is further configured to predict the future humidity level and / or calculate the desired temperature setpoint (SP2) of the secondary control module based on sensor data related to parameters affecting the humidity level being controlled.
5. The MPC module: - a prediction module (222, 322), and - optimization module (224, 324), Including, The prediction module (222, 322) is configured to predict the future humidity level, and predicting the future humidity level comprises: - generating a number of successive future humidity levels; Generating a plurality of future desired temperature setpoints; further comprising 5. The cascade control system of claim 1, wherein the optimization module (224, 324) is configured to calculate the desired temperature setpoint of the secondary control module, and the calculation of the desired temperature setpoint is based on at least one of a humidity level control performance metric using the plurality of consecutive future humidity levels, a desired temperature setpoint stability metric using the plurality of future desired temperature setpoints, and an energy efficiency performance metric using the plurality of future desired temperature setpoints.
6. 6. A cascade control system according to claim 1, wherein the prediction of future humidity levels is performed using a linear model.
7. The cascade control system of claim 6 , wherein the desired temperature setpoint calculation is determined from a closed form solution.
8. 8. The cascade control system of claim 1, further comprising a primary PID control module (240, 340) configured to control the humidity level by controlling the at least one operating parameter of the dehumidifier and to provide a desired temperature setpoint (SP2) for the secondary control module when predetermined criteria are met.
9. 9. The cascade control system of claim 1, further configured to adjust the desired temperature setpoint (SP2) of the secondary control module based on the temperature and / or humidity level of the effluent reactivation air.
10. a tertiary control module (250, 350) for controlling the temperature of the outflow reactivation air by adjusting the desired temperature setpoint (SP2) of the secondary control module based on the difference between the temperature of the outflow reactivation air and a predetermined threshold temperature; The cascade control system of claim 1 , further comprising:
11. a quaternary control module (260, 360), receiving a desired room humidity setpoint (SP3) related to the humidity level of the room (270, 370) to which said effluent process air is to be supplied; - receiving the current humidity level in the room; Generate the desired humidity set point (SP1) for the primary control module; a quaternary control module configured to: The cascade control system of claim 1 , further comprising:
12. A dehumidification system (3000) for controlling humidity levels, comprising: a dehumidifier (302), and A cascade control system (200, 300) according to any one of claims 1 to 11, Dehumidification system with.