Data-driven Control Strategy Research Articles

Utilizing a optimized method for evaluating vapor recovery equipment control efficiency and estimating evaporative VOC emissions from urban oil depots via an extensive survey

As an important intermediary between upstream refineries and downstream urban gas stations, volatile organic compound (VOC) emissions from urban oil depots were often disregarded, underestimating their environmental and health implications. An extensive investigation of urban depots' fuel composition and operational dynamics was conducted nationwide. We developed a novel approach that integrates theoretical models with easily measurable operational data from the depots to evaluate the efficiency of post-treatment devices in actual situations. Even in well-managed oil depots, the actual control efficiency of vapor recovery units fluctuates between 63 % and 85 %, depending on the concentration of hydrocarbon vapors in the intake of the device. The national emission factors for gasoline, diesel, and aviation kerosene at a national level were 6.64 ± 1.16, 2.07 ± 0.42, and 6.17 ± 1.05 tons per 10,000 tons, respectively. In 2019, China's urban oil depots emitted 165 thousand tons of VOC. Enhancing control strategies by optimizing the physical and chemical parameters of refined oil, improving storage capacity and turnover efficiency, and upgrading storage tanks had the potential to reduce emissions by more than 60 %. However, a 30 % failure rate in these systems could negate the benefits of these improved strategies.

Unified control of diverse actions in a wastewater treatment activated sludge system using reinforcement learning for multi-objective optimization

The operation of modern wastewater treatment facilities is a balancing act in which a multitude of variables are controlled to achieve a wide range of objectives, many of which are conflicting. This is especially true within secondary activated sludge systems, where significant research and industry effort has been devoted to advance control optimization strategies, both domain-driven and data-driven. Among data-driven control strategies, reinforcement learning (RL) stands out for its ability to achieve better than human performance in complex environments. While RL has been applied to activated sludge process optimization in existing literature, these applications are typically limited in scope, and never for the control of more than three actions. Expanding the scope of RL control has the potential to increase the optimization potential while concurrently reducing the number of control systems that must be tuned and maintained by operations staff. This study examined several facets of the implementation of multi-action, multi-objective RL agents, namely how many actions a single agent could successfully control and what extent of environment data was necessary to train such agents. This study observed improved control optimization with increasing action scope, though control of waste activated sludge remains a challenge. Furthermore, agents were able to maintain a high level of performance under decreased observation scope, up to a point. When compared to baseline control of the Benchmark Simulation Model No. 1 (BSM1), an RL agent controlling seven individual actions improved the average BSM1 performance metric by 8.3 %, equivalent to an annual cost savings of $40,200 after accounting for the cost of additional sensors.

Koopman-Based MPC With Learned Dynamics: Hierarchical Neural Network Approach.

This article presents a data-driven control strategy for nonlinear dynamical systems, enabling the construction of a Koopman-based linear system associated with nonlinear dynamics. The primary idea is to apply the deep learning technique to the Koopman framework for globally linearizing nonlinear dynamics and impose a Koopman-based model predictive control (MPC) approach to stabilize the nonlinear dynamical systems. In this work, we first generalize the Koopman framework to nonlinear control systems, enabling comprehensive linear analysis and control methods to be effective for nonlinear systems. We next present a hierarchical neural network (HNN) approach to deal with the crucial challenge of the finite-dimensional Koopman representation approximation. In particular, a scale-invariant constrained network in the HNN includes four modules, in which a predictor module and a linear module can accurately approximate the finite Koopman eigenfunctions and Koopman operator, respectively, thus forming the lifted linear system. Then, we design the Koopman-based MPC scheme for controlling nonlinear systems with constraints by adopting the modified MPC with a saturation-like function on the lifted linear system. Importantly, the Koopman-based MPC enjoys higher computational efficiency compared to the classical linear MPC and nonlinear MPC methods. Finally, a physical experiment on an overhead crane system is provided to demonstrate the effectiveness of the proposed data-driven control framework.

Enhanced fault detection in polymer electrolyte fuel cells via integral analysis and machine learning

The growing energy demand and population raising require alternative, clean, and sustainable energy systems. During the last few years, hydrogen energy has proven to be a crucial factor under the current conditions. Although the energy conversion process in polymer electrolyte fuel cells (PEFCs) is clean and noiseless since the only by-products are heat and water, the inside phenomena are not simple. As a result, correct monitoring of the health situation of the device is required to perform efficiently. This paper aims to explore and evaluate the machine learning (ML) and deep learning (DL) models for predicting classification fault detection in PEFCs. It represents a support for decision-making by the fuel cell operator or user. Seven ML and DL model classifiers are considered. A database comprising 182,156 records and 20 variables arising from the fuel cell's energy conversion process and operating conditions is considered. This dataset is unbalanced; therefore, techniques to balance are applied and analyzed in the training and testing of several models. The results showed that the logistic regression (LR), k-nearest neighbor (KNN), decision tree (DT), random forest (RF), and Naive Bayes (NB) models present similar and optimal trends in terms of performance indicators and computational cost; unlike support vector machine (SMV) and multi-layer perceptron (MLP) whose performance is affected when the data is balanced and even presents a higher computational cost. Therefore, it is a novel approach for fault detection analysis in PEFC that combines the interpretability of different ML and DL algorithms while addressing data imbalance, so common in the real world, using resampling techniques. This methodology provides clear information for the model decision-making process, improving confidence and facilitating further optimization; in contrast to traditional physics-based models, paving the way for data-driven control strategies.

Feedforward-feedback-enhanced model-free adaptive iterative learning control with measurement disturbance and data dropout for an autonomous bus trajectory tracking system.

This article presents an innovative enhanced model-free adaptive iterative learning control approach suited for autonomous bus trajectory tracking systems that may experience measurement disruptions and random data dropouts. Data loss can occur independently and randomly at different times and in different iterations with varying probabilities, leading to successive data dropouts on both the time and iteration axes. The proposed enhanced model-free adaptive iterative learning control controller incorporates a data compensation mechanism to compensate for missing data, ensuring excellent control performance. This data-driven control strategy requires only input/output data for controller design. The convergence and effectiveness of the proposed approach are verified through rigorous mathematical analysis and simulation outcomes.

Data-driven control of wave energy systems using random forests and deep neural networks

In the field of sustainable energy and alternatives to fossil fuels, wave energy is generating an increasing interest due to its untapped potential. However, the levelised cost of energy of wave energy systems is still not able to compete with other renewable technologies, mainly due to high costs associated with their conversion process. In this context, the development of energy-maximising control strategies plays an important role towards the economic viability of wave energy technology, by optimising the overall energy conversion, hence contributing towards minimising the associated cost of energy. State-of-the-art control systems for wave energy converters are mostly model-based, exploiting control-oriented models of the device to compute the applied control actions with a limited computational burden. Nevertheless, these representations of the system are simplified, and often based upon unrealistic assumptions, such as small motion around the zero equilibrium position, which are inherently invalidated during device operations and that can lead to large uncertainties, resulting in suboptimal power absorption. For these reasons, in this paper, a purely data-driven control strategy is developed. This strategy exploits random forests (RFs) and deep neural networks (DNNs) to gradually learn from real experiences towards an optimal proportional–integral control action. These structures are used as surrogate models (built upon the data coming from past experiences) to converge to the optimal control parameters in a surrogate-optimisation-like manner. To manage the exploration and exploitation needs of controllers based on this approach, a learning strategy is developed and presented. Some considerations are made on the choice of the input features of the surrogate structures, which deeply affect the control strategy learning results. To assess the performances of both the control and learning strategies, one year of operations has been simulated under control settings guided by the proposed data-driven approach, showing also the potential capabilities that the adoption of RFs and DNNs has in learning, even in sea conditions with a limited number of occurrences.

Data-Driven Bipartite Consensus Tracking for Nonlinear Multiagent Systems With Prescribed Performance

In this article, a data-driven control strategy with prescribed performance is proposed to address the distributed bipartite consensus problem for unknown nonlinear multiagent systems (MASs) undersigned directed graphs. The dynamics of the agents in each topology are completely unknown and the desired trajectory information is only communicated to a subset of agents. First, the unknown nonlinear dynamic of each agent is transformed into an equivalent time-varying linearized model by utilizing the dynamic linearization method. Then, by considering the prescribed performance, a strictly increasing function is given to transform the constrained tracking error condition of the nonlinear MASs into an equivalent unconstrained error condition. Meanwhile, a new control protocol that only uses the input and output data of the agents is designed in the case of fixed or switched topologies. The proposed method can not only guarantee the bipartite consensus but also ensure the bipartite tracking error always converges to the prescribed region. Moreover, the designed control parameters can be adjusted appropriately to obtain better tracking performance. Finally, the convergence of bipartite consensus tracking error is guaranteed through rigorous theoretical analysis. The control scheme is further verified by two examples.

Formation Control of Non-Holonomic Mobile Robots: Predictive Data-Driven Fuzzy Compensator

A key research topic in the field of robotics is the formation control of a group of robots in trajectory tracking problems. Using organized robots has many advantages over using them individually, such as efficient use of resources, increased reliability due to cooperation, and better resistance against defects. To achieve this, a controller is proposed that steers the leader robot and subsequent follower robots asymptotically to a reference trajectory. The basic controller is feedback linearization. To ensure stability against perturbations, a compensator based on type-3 fuzzy logic systems (T3-FLSs) and a data-driven control strategy is designed. The approach involves employing a finite number of open-loop data and using the model-based predictive controller (MPC) approach to acquire sufficient criteria for stability. An infinite-horizon function is minimized online, which allows the data-based control policy to be considered the optimal control method. The gains of the constrained data-based control signal are computed at each time step to enhance accuracy. Applying the data-based state feedback controller to the system yields positive and stable state trajectories with appropriate transient responses. The suggested data-driven compensator is guaranteed to handle constraints. A practical example is simulated to evaluate the proposed strategy.

Deep reinforcement learning for turbulent drag reduction in channel flows

We introduce a reinforcement learning (RL) environment to design and benchmark control strategies aimed at reducing drag in turbulent fluid flows enclosed in a channel. The environment provides a framework for computationally efficient, parallelized, high-fidelity fluid simulations, ready to interface with established RL agent programming interfaces. This allows for both testing existing deep reinforcement learning (DRL) algorithms against a challenging task, and advancing our knowledge of a complex, turbulent physical system that has been a major topic of research for over two centuries, and remains, even today, the subject of many unanswered questions. The control is applied in the form of blowing and suction at the wall, while the observable state is configurable, allowing to choose different variables such as velocity and pressure, in different locations of the domain. Given the complex nonlinear nature of turbulent flows, the control strategies proposed so far in the literature are physically grounded, but too simple. DRL, by contrast, enables leveraging the high-dimensional data that can be sampled from flow simulations to design advanced control strategies. In an effort to establish a benchmark for testing data-driven control strategies, we compare opposition control, a state-of-the-art turbulence-control strategy from the literature, and a commonly used DRL algorithm, deep deterministic policy gradient. Our results show that DRL leads to 43% and 30% drag reduction in a minimal and a larger channel (at a friction Reynolds number of 180), respectively, outperforming the classical opposition control by around 20 and 10 percentage points, respectively.

A Data-Driven and Personalized Stance Symmetry Controller for Robotic Ankle-Foot Prostheses: A Preliminary Investigation.

People with unilateral transtibial amputation generally exhibit asymmetric gait, likely due to inadequate prosthetic ankle function. This results in compensatory behavior, leading to long-term musculoskeletal impairments (e.g., osteoarthritis in the joints of the intact limb). Powered prostheses can better emulate biological ankles, however, control methods are over-reliant on able-bodied data, require extensive amounts of tuning by experts, and cannot adapt to each user's unique gait patterns. This work directly addresses all these limitations with a personalized and data-driven control strategy. Our controller uses a virtual setpoint trajectory within an impedance-inspired formula to adjust the dynamics of the robotic ankle-foot prosthesis as a function of stance phase. A single sensor measuring thigh motion is used to estimate the gait phase in real time. The virtual setpoint trajectory is modified via a data-driven iterative learning strategy aimed at optimizing ankle angle symmetry. The controller was experimentally evaluated on two people with transtibial amputation. The control scheme successfully increased ankle angle symmetry about the two limbs by 24.4% when compared to the passive condition. In addition, the symmetry controller significantly increased peak prosthetic ankle power output at push-off by 0.52 W/kg and significantly reduced biomechanical risk factors associated with osteoarthritis (i.e., knee and hip abduction moments) in the intact limb. This research demonstrates the benefits of personalized and data-driven symmetry controllers for robotic ankle-foot prostheses.

Retracted: A Data-Driven Control Strategy for Urban Express Ramp.

[This retracts the article DOI: 10.1155/2021/6540972.].

Real-Time Data-Driven Approach for Prediction and Correction of Electrode Array Trajectory in Cochlear Implantation

Cochlear implants provide hearing perception to people with severe to profound hearing loss. The electrode array (EA) inserted during the surgery directly stimulates the hearing nerve, bypassing the acoustic hearing system. The complications during the EA insertion in the inner ear may cause trauma leading to infection, residual hearing loss, and poor speech perception. This work aims to reduce the trauma induced during electrode array insertion process by carefully designing a sensing method, an actuation system, and data-driven control strategy to guide electrode array in scala tympani. Due to limited intra-operative feedback during the insertion process, complex bipolar electrical impedance is used as a sensing element to guide EA in real time. An automated actuation system with three degrees of freedom was used along with a complex impedance meter to record impedance of consecutive electrodes. Prediction of EA direction (medial, middle, and lateral) was carried out by an ensemble of random forest, shallow neural network, and k-nearest neighbour in an offline setting with an accuracy of 86.86%. The trained ensemble was then utilized in vitro for prediction and correction of EA direction in real time in the straight path with an accuracy of 80%. Such a real-time system also has application in other electrode implants and needle and catheter insertion guidance.

Learning of model-plant mismatch map via neural network modeling and its application to offset-free model predictive control

We propose an improved offset-free model predictive control (MPC) framework, which learns and utilizes the intrinsic model-plant mismatch map, to effectively exploit the advantages of model-based and data-driven control strategies and overcome the limitation of each approach. In this study, the model-plant mismatch map on steady-state manifold is approximated via artificial neural network (ANN) modeling based on steady-state data from the process. Though the learned model-plant mismatch map can provide the information at the equilibrium point (i.e., setpoint), it cannot provide model-plant mismatch information during transient state. To handle this, we additionally apply a supplementary disturbance variable which is updated from a revised disturbance estimator considering the disturbance value obtained from the learned model-plant mismatch map. Then, the learned and supplementary disturbance variables are applied to the target problem and finite-horizon optimal control problem of the offset-free MPC framework. By this, the control system can utilize both the learned model-plant mismatch information and the stabilizing property of the nominal disturbance estimator. The closed-loop simulation results demonstrate that the proposed offset-free MPC scheme utilizing the model-plant mismatch map learned via ANN modeling efficiently improves the closed-loop reference tracking performance of the control system. Additionally, the zero-offset tracking condition of the developed framework is mathematically examined.

Input torque measurements for wind turbine gearboxes using fiber-optic strain sensors

Abstract. Accurate knowledge of the input torque in wind turbine gearboxes is key to improving their reliability. Traditionally, rotor torque is measured using strain gauges bonded to the shaft. Transferring the resulting signal from the rotating shaft to a stationary data acquisition system while powering the sensing devices is complex and costly. The magnitude of the torques involved in wind turbine gearboxes and the high stiffness of the input shaft pose additional difficulties. This paper presents a new alternative method to measure the input torque in wind turbine gearboxes based on deformation measurements of the static first-stage ring gear. We have measured deformation using fiber-optic strain sensors based on fiber Bragg gratings because of their advantages compared to conventional electrical strain gauges. The present study was conducted on a Siemens Gamesa Renewable Energy gearbox with a rated power of 6 MW, in which a total of 54 fiber-optic strain sensors were installed on the outer surface of the first-stage ring gear. The gear mesh forces between the planets and the ring gear cause measurable deformations on the outer surface of the stationary ring gear. The measured strains exhibit a dynamic behavior. The strain values change depending on the relative position of the strain sensors to the planet gears, the instantaneous variations of the input torque, and the way load is shared between planets. A satisfactory correlation has been found between the strain signals measured on the static ring gear and torque. Two signal processing strategies are presented in this paper. The first procedure is based on the peak-to-peak strain values computed for the gear mesh events, and therefore torque can only be estimated when a gear mesh event is detected. The second signal processing procedure combines the strain signals from different sensors using a Coleman coordinate transformation and tracks the magnitude of the fifth harmonic component. With this second procedure, it is possible to estimate torque whenever strain data of all sensors are available, leading to an improved frequency resolution up to the sampling frequency used to acquire strain data. The method presented in this paper could make measuring gearbox torque more cost-effective, which would facilitate its adoption in serial wind turbines and enable novel data-driven control strategies, as well as a more accurate assessment of the consumed fatigue life of the gearboxes throughout their operation.

Improved offset-free model predictive control utilizing learned model-plant mismatch map

The requirement for a framework that effectively overcomes the limitation of model-based and data-driven control strategies by combining both methods continues to grow. In this study, we propose an approach that learns the model-plant mismatch map and utilizes it based on the offset-free model predictive control (MPC). Specifically, the mismatch map is learned via general regression neural network (GRNN) that has been applied in broad range of fields based on the data from the process, and then the learned mismatch information is provided to the MPC system. In addition, since the approximated mismatch map via GRNN cannot be perfect, an additional supplementary disturbance estimator is utilize to ensure the zero-offset tracking property. Finally, the learned and supplementary disturbance signals are applied to the target problem and the optimal control problem based on the offset-free MPC framework. The effectiveness of the proposed combined model-based and data driven framework is demonstrated by closed-loop simulation. The result shows that the proposed framework can improve the closed-loop tracking performance by utilizing both the learned mismatch information from GRNN and the stabilizing property of the supplementary disturbance estimator.

Data-driven model identification and predictive control for path-following of underactuated ships with unknown dynamics

With the trends towards autonomous shipping, advanced ship motion control methods have received increased attention in recent years. The validity of ship models is crucial in designing motion controllers and directly affects their performances. However, accurate models that could reflect true ship dynamics are highly nonlinear, complex and complicated to identify, especially in situations when the experimental conditions are limited. This paper proposes a data-driven predictive control method for path-following of under-actuated cargo ships with unknown dynamics, which makes use of data gathered during operation to improve the model and the path-following performance. Based on the ship navigation data set, the relations between the heading angle and the rudder angle of the ship are fitted with seven typical regression algorithms, which acts as the prediction model in the path-following controller. Simulation study is carried out to choose the most suitable regression algorithm, among which elastic net regression is selected. The Antenna Mutation Beetle Swarm Predictive (AMBS-P) algorithm is introduced to find the optimal weights in the model identification process. A Line-of-Sight (LOS) algorithm is used as the guidance law to transform reference way-points into reference heading angles, and the path-following controller is designed also based on and the AMBS-P algorithm. Simulation results show that the proposed data-driven control method performs well in the path-following task without having prior knowledge regarding the hydrodynamic coefficients and ship parameters. • The proposed data-driven control strategy allows the users to exploit the advantage of the predictive control scheme, as well as exploiting the approximate model knowledge captured from data without having prior knowledge regarding the hydrodynamic coefficients and ship parameters. • A quantitative analysis of the model identification performances of seven typical regression algorithms is carried out. • The Antenna Mutation Beetle Swarm Predictive (AMBS-P) algorithm is proposed and applied in finding the optimal weights in the model identification process and in designing the path-following controller with a LOS guidance principle. This could save the efforts spend in weight selection and controller parameter adjustments. It is also fast and easy to implement, which could facilitate its use in practice.

Data Driven Feedforward Control Strategy for Multi-Fuel UAS Engine

An engine is a complex system that requires a control strategy to perform optimally and reliably. The majority of the existing system uses a lookup table-based control strategy, also known as feedforward tables, generated offline by performing engine calibration, which is usually an expensive process. The calibration cost can be significantly reduced using a system model, but due to the increased engine complexity, developing a physics-based model that can capture all the system modes becomes challenging. Therefore, the current work uses a data-driven approach to model a diesel engine and designs its control strategy based on it. It uses Gaussian Process Regression (GPR) to model the engine and perform system inversion to achieve the desired optimal feedforward control. Two control inputs, namely, pilot injection timing (PIT) and main injection timing (MIT), are calculated to achieve the desired combustion performance in terms of CA50 from the engine. Model inversion is done using the real-variable genetic algorithm (rGA) to obtain the optimal control strategy corresponding to the desired CA50. The study tries to address two fundamental issues for developing a data-driven control strategy for a practical system: when the model is developed using limited number of data points, and the inversion problem when multiple control settings generate the same optimal output. Finally, the proposed control strategy is validated on an actual experimental test bench, and the results show excellent performance in achieving the desired CA50 value.

Data-Driven Aggregation Control for Thermoelectric Loads in Demand Response

Within the concept of a smart grid, aggregators have the task of coordinating the behavior of large sets of Distributed Energy Resources, each of them offering small power/energy capacities, which help to balance the power grid and can serve as providers of services. Adequate coordination strategies are required to optimally exploit these resources in the ancillary services market. However, deriving model-based control policies for them is complex due to the heterogeneity and uncertainty related to the large set of associated agents. Then, a data-driven model is an adequate solution for this sort of situation. This paper presents the application of the Youla–Kucera Data-Driven Control strategy for the development of an aggregator to regulate the power consumption of a set of thermoelectric refrigerators, avoiding the modeling process and directly designing a controller from data. A detailed simulation framework was executed to verify the validity of the proposed methodology. It is shown that the derived aggregator is able to offer frequency containment reserves service, achieving the required settling time of 30 seconds and with a tracking error below 4.7%.

Input Torque Measurements for Wind Turbine Gearboxes Using Fiber Optical Strain Sensors

Abstract. Accurate knowledge of the input torque in wind turbine gearboxes is key to improving their reliability. Traditionally, rotor torque is measured using strain gauges bonded to the shaft. Transferring the resulting signal from the rotating shaft to a stationary data acquisition system while powering the sensing devices is complex and costly. The magnitude of the torques involved in wind turbine gearboxes and the high stiffness of the input shaft pose additional difficulties. This paper presents a new alternative method to measure the input torque in wind turbine gearboxes based on deformation measurements of the static first stage ring gear. We have measured deformation using fiber optic strain sensors based on fiber Bragg gratings because of their advantages compared to conventional electrical strain gauges. The present study was conducted on a Siemens Gamesa Renewable Energy gearbox with a rated power of 6MW, in which a total of 54 fiber optic strain sensors were installed on the outer surface of the first stage ring gear. The gear mesh forces between the planets and the ring gear cause measurable deformations on the outer surface of the stationary ring gear. The measured strains exhibit a dynamic behavior. The strain values change depending on the relative position of the strain sensors to the planet gears, the instantaneous variations of the input torque, and the way load is shared between planets. A satisfactory correlation has been found between the strain signals measured on the static ring gear and torque. Two signal processing strategies are presented in this paper. The first procedure is based on the peak-to-peak strain values computed for the gear mesh events, and therefore, torque can only be estimated when a gear mesh event is detected. The second signal processing procedure combines the strain signals from different sensors using a Coleman coordinate transformation and tracks the magnitude of the fifth harmonic component. With this second procedure, it is possible to estimate torque whenever strain data of all sensors is available, leading to an improved frequency resolution up to the sampling frequency used to acquire strain data. The method presented in this paper could make measuring gearbox torque more cost-effective, which would facilitate its adoption in serial wind turbines and enable novel data-driven control strategies, as well as a more accurate assessment of the consumed fatigue life of the gearboxes throughout their operation.

Exploring New Building Energy Saving Control Strategy Application under the Energy Internet of Things

This paper makes innovative research on developing a data-driven control strategy under the Energy Internet of Things architecture. On the one hand, the platform aims to provide data representation and interpretable analysis for different stakeholders (end users, construction operators or managers) to realize the flexibility and scalability of the platform; on the other hand, it can improve the thermal comfort and also reduce the power consumption of buildings. However, to process vast amounts of data, it is critical to select appropriate control methods and design optimization issues. Data-driven predictive control (DPC) is a control technology that replaces model-based predictive control (MPC). When applied to complex building operations, MPC is implemented by using the control-oriented data-driven model. The key to DPC technology is the use of CatBoost algorithm, which is highly interpretable and easy to be operated by stakeholders. This paper chooses TDNN, LightGBM, and CatBoost to compare and analyze building energy consumption. Numerical simulation results show that the CatBoost algorithm’s performance is better than other algorithms, and the complexity and implementation cost is significantly reduced.