One major issue in learning-based model predictive control (MPC) for autonomous driving is the contradiction between the system model's prediction accuracy and computation efficiency. The more situations a system model covers, the more complex it is, along with highly nonlinear and nonconvex properties. These issues make the optimization too complicated to solve and render real-time control impractical.To address these issues, we propose a hierarchical learning residual model which leverages random forests and linear regression.The learned model consists of two levels. The low level uses linear regression to fit the residues, and the high level uses random forests to switch different linear models. Meanwhile, we adopt the linear dynamic bicycle model with error states as the nominal model.The switched linear regression model is added to the nominal model to form the system model. It reformulates the learning-based MPC as a quadratic program (QP) problem and optimization solvers can effectively solve it. Experimental path tracking results show that the driving vehicle's prediction accuracy and tracking accuracy are significantly improved compared with the nominal MPC.Compared with the state-of-the-art Gaussian process-based nonlinear model predictive control (GP-NMPC), our method gets better performance on tracking accuracy while maintaining a lower computation consumption.

Recent surge in the number of Electric Vehicles have created a need to develop inexpensive energy-dense Battery Storage Systems. Many countries across the planet have put in place concrete measures to reduce and subsequently limit the number of vehicles powered by fossil fuels. Lithium-ion based batteries are presently dominating the electric automotive sector. Energy research efforts are also focussed on accurate computation of State-of-Charge of such batteries to provide reliable vehicle range estimates. Although such estimation algorithms provide precise estimates, all such techniques available in literature presume availability of superior quality battery datasets. In reality, gaining access to proprietary battery usage datasets is very tough for battery scientists. Moreover, open access datasets lack the diverse battery charge/discharge patterns needed to build generalized models. Curating battery measurement data is time consuming and needs expensive equipment. To surmount such limited data scenarios, we introduce few Deep Learning-based methods to synthesize high-fidelity battery datasets, these augmented synthetic datasets will help battery researchers build better estimation models in the presence of limited data. We have released the code and dataset used in the present approach to generate synthetic data. The battery data augmentation techniques introduced here will alleviate limited battery dataset challenges.

Accurate time series forecasting is critical for a wide range of problems with temporal data. Ensemble modeling is a well-established technique for leveraging multiple predictive models to increase accuracy and robustness, as the performance of a single predictor can be highly variable due to shifts in the underlying data distribution. This paper proposes a new methodology for building robust ensembles of time series forecasting models. Our approach utilizes Adaptive Robust Optimization (ARO) to construct a linear regression ensemble in which the models' weights can adapt over time. We demonstrate the effectiveness of our method through a series of synthetic experiments and real-world applications, including air pollution management, energy consumption forecasting, and tropical cyclone intensity forecasting. Our results show that our adaptive ensembles outperform the best ensemble member in hindsight by 16-26% in root mean square error and 14-28% in conditional value at risk and improve over competitive ensemble techniques.

Multivariate time series (MTS) forecasting plays an important role in the automation and optimization of intelligent applications. It is a challenging task, as we need to consider both complex intra-variable dependencies and inter-variable dependencies. Existing works only learn temporal patterns with the help of single inter-variable dependencies. However, there are multi-scale temporal patterns in many real-world MTS. Single inter-variable dependencies make the model prefer to learn one type of prominent and shared temporal patterns. In this paper, we propose a multi-scale adaptive graph neural network (MAGNN) to address the above issue. MAGNN exploits a multi-scale pyramid network to preserve the underlying temporal dependencies at different time scales. Since the inter-variable dependencies may be different under distinct time scales, an adaptive graph learning module is designed to infer the scale-specific inter-variable dependencies without pre-defined priors. Given the multi-scale feature representations and scale-specific inter-variable dependencies, a multi-scale temporal graph neural network is introduced to jointly model intra-variable dependencies and inter-variable dependencies. After that, we develop a scale-wise fusion module to effectively promote the collaboration across different time scales, and automatically capture the importance of contributed temporal patterns. Experiments on four real-world datasets demonstrate that MAGNN outperforms the state-of-the-art methods across various settings.

The Common Information (CI) approach provides a systematic way to transform a multi-agent stochastic control problem to a single-agent partially observed Markov decision problem (POMDP) called the coordinator's POMDP. However, such a POMDP can be hard to solve due to its extraordinarily large action space. We propose a new algorithm for multi-agent stochastic control problems, called coordinator's heuristic search value iteration (CHSVI), that combines the CI approach and point-based POMDP algorithms for large action spaces. We demonstrate the algorithm through optimally solving several benchmark problems.

Much of the conversation around artificial intelligence (AI) these days centers on whether it will eventually take your job, how it's trying to compete with humans in creative fields, or how it can be misused, say, as a writing tool. You can probably chalk this one-sidedness up to an all-too-human tendency to be suspicious of new tech that isn't well understood by the mainstream (yet). But AI isn't intrinsically evil or good: It's a tool, a vast technology with enormous potential, and there are myriad ways to implement it beyond the current discourse. One vitally important use case is helping us fight and survive the consequences of climate change. Whether it's mitigating the effects of disasters such as floods and fires more quickly or building a cleaner energy grid, the evidence is mounting that AI has an essential role to play in helping to protect us as the planet reacts to climate change.

Incumbents like CBA are increasingly looking to A.I. technology to solve their business problems and are eyeing external tech partners to source those A.I. solutions. But these traditional companies have faced challenges nurturing meaningful collaborations that maximize the support they get from A.I. players. Only 1 in 5 incumbents found the right kind of A.I. player, like H20.ai is for CBA, that offers access to custom technology, as well as support for talent, training, and change management, prompting the incumbent to overhaul its processes. We call these A.I. players that provide such support transformers. For industry incumbents that are able to identify and effectively collaborate with a transformer, the value is clear.

Vattenfall is a European energy company with approximately 20 000 employees. For more than 100 years we have electrified industries, supplied energy to people's homes and modernized our way of living through innovation and cooperation. We now want to make fossil-free living possible within one generation. To be able to reach this ambitious goal we are looking for talented individuals who, in addition to their passion for their own role, also have strong team spirit and want to contribute to supporting a meaningful corporate mission. We offer power purchase agreements to operators of wind or solar parks, optimize 3rd party batteries and enter into origination deals within the B2B segment.

Sampling from high-dimensional distributions is a fundamental problem in statistical research and practice. However, great challenges emerge when the target density function is unnormalized and contains isolated modes. We tackle this difficulty by fitting an invertible transformation mapping, called a transport map, between a reference probability measure and the target distribution, so that sampling from the target distribution can be achieved by pushing forward a reference sample through the transport map. We theoretically analyze the limitations of existing transport-based sampling methods using the Wasserstein gradient flow theory, and propose a new method called TemperFlow that addresses the multimodality issue. TemperFlow adaptively learns a sequence of tempered distributions to progressively approach the target distribution, and we prove that it overcomes the limitations of existing methods. Various experiments demonstrate the superior performance of this novel sampler compared to traditional methods, and we show its applications in modern deep learning tasks such as image generation. The programming code for the numerical experiments is available at https://github.com/yixuan/temperflow.

Overloading in DC servo motors is a major concern in industries, as many companies face the problem of finding expert operators, and also human monitoring may not be an effective solution. Therefore, this paper proposed an embedded Artificial intelligence (AI) approach using a Convolutional Neural Network (CNN) using a new transformation to extract faults from real-time input signals without human interference. Our main purpose is to extract as many as possible features from the input signal to achieve a relaxed dataset that results in an effective but compact network to provide real-time fault detection even in a low-memory microcontroller. Besides, fault detection method a synchronous dual-motor system is also proposed to take action in faulty events. To fulfill this intention, a one-dimensional input signal from the output current of each DC servo motor is monitored and transformed into a 3d stack of data and then the CNN is implemented into the processor to detect any fault corresponding to overloading, finally experimental setup results in 99.9997% accuracy during testing for a model with nearly 8000 parameters. In addition, the proposed dual-motor system could achieve overload reduction and provide a fault-tolerant system and it is shown that this system also takes advantage of less energy consumption.