Mogens Graf Plessen

Mogens Graf Plessen

A method for harvest planning based on the coupling of crop assignment with vehicle routing is presented. Given a setting with multiple fields, a path network connecting these, multiple depots at which a number of harvesters are initially located, the main question addressed is: Which crop out of a set of different crops to assign to each field when accounting for the given setting? It must be answered by every farm manager at the beginning of every work-cycle starting with plant seeding and ending with harvesting. Rather than solving an assignment problem only, it is here also accounted for the connectivity between fields. In practice, fields are often located distant apart. Traveling costs of machinery and limited harvesting windows demand optimised operation and route planning. Therefore, the proposed method outputs crop assignments to fields and simultaneously determines crop-tours, i.e., optimised sequences in which to service fields of the same crop during harvest. It is of particular relevance for larger farms and groups of farms that collaborate and share machinery. Integer programming based exact algorithms are derived. For large numbers of fields, where these algorithms may not be tractable due to computational constraints, elements of clustering and the solution of local Traveling Salesman Problems are added, thereby on the one hand rendering the method heuristic and in general suboptimal, but on the other hand maintaining large-scale applicability.

Mogens Graf Plessen

This paper presents a path planning method for partial field coverage. Therefore, a specific path planning pattern is proposed. The notion is that lighter machinery with smaller storage tanks can alleviate soil compaction because of a reduced weight, but does not enable full field coverage in a single run because of the smaller storage capacity. This is relevant for spraying applications and related in-field work. Consequently, multiple returns to a mobile or stationary depot located outside of the field are required for storage tank refilling. Therefore, a suitable path planning method is suggested that accounts for the limited turning radii of agricultural vehicles, satisfies compacted area minimisation constraints, and aims at overall path length minimisation. The benefits of the proposed method are illustrated by means of a comparison to a planning method based on the more common AB pattern. It is illustrated how the proposed path planning pattern can also be employed efficiently for single-run field coverage.

Mogens Graf Plessen

This paper proposes a simple method to extract from a set of multiple related time series a compressed representation for each time series based on statistics for the entire set of all time series. This is achieved by a hierarchical algorithm that first generates an alphabet of shapelets based on the segmentation of centroids for clustered data, before labels of these shapelets are assigned to the segmentation of each single time series via nearest neighbor search using unconstrained dynamic time warping as distance measure to deal with non-uniform time series lenghts. Thereby, a sequence of labels is assigned for each time series. Completion of the last label sequence permits prediction of individual time series. Proposed method is evaluated on two global COVID-19 datasets, first, for the number of daily net cases (daily new infections minus daily recoveries), and, second, for the number of daily deaths attributed to COVID-19 as of April 27, 2020. The first dataset involves 249 time series for different countries, each of length 96. The second dataset involves 264 time series, each of length 96. Based on detected anomalies in available data a decentralized exit strategy from lockdowns is advocated.

Mogens Graf Plessen

Within the context of multivariate time series segmentation this paper proposes a method inspired by a posteriori optimal trading. After a normalization step time series are treated channel-wise as surrogate stock prices that can be traded optimally a posteriori in a virtual portfolio holding either stock or cash. Linear transaction costs are interpreted as hyperparameters for noise filtering. Resulting trading signals as well as resulting trading signals obtained on the reversed time series are used for unsupervised labeling, before a consensus over channels is reached that determines segmentation time instants. The method is model-free such that no model prescriptions for segments are made. Benefits of proposed approach include simplicity, adaptability to a wide range of different shapes of time series, and in particular computational efficiency that make it suitable for big data. Performance is demonstrated on synthetic and real-world data, including a large-scale dataset comprising a multivariate time series of dimension 1000 and length 2709. Proposed method is compared to a popular model-based bottom-up approach fitting piecewise affine models and to a state-of-the-art model-based top-down approach fitting Gaussian models, and found to be consistently faster while producing more intuitive results.

Mogens Graf Plessen

This paper proposes online sampling in the parameter space of a neural network for GPU-accelerated motion planning of autonomous vehicles. Neural networks are used as controller parametrization since they can handle nonlinear non-convex systems and their complexity does not scale with prediction horizon length. Network parametrizations are sampled at each sampling time and then held constant throughout the prediction horizon. Controls still vary over the prediction horizon due to varying feature vectors fed to the network. Full-dimensional vehicles are modeled by polytopes. Under the assumption of obstacle point data, and their extrapolation over a prediction horizon under constant velocity assumption, collision avoidance reduces to linear inequality checks. Steering and longitudinal acceleration controls are determined simultaneously. The proposed method is designed for parallelization and therefore well-suited to benefit from continuing advancements in hardware such as GPUs. Characteristics of proposed method are illustrated in 5 numerical simulation experiments including dynamic obstacle avoidance, waypoint tracking requiring alternating forward and reverse driving with maximal steering, and a reverse parking scenario.

Mogens Graf Plessen

Within the context of trajectory planning for autonomous vehicles this paper proposes methods for efficient encoding of motion primitives in neural networks on top of model-based and gradient-free reinforcement learning. It is distinguished between 5 core aspects: system model, network architecture, training algorithm, training tasks selection and hardware/software implementation. For the system model, a kinematic (3-states-2-controls) and a dynamic (16-states-2-controls) vehicle model are compared. For the network architecture, 3 feedforward structures are compared including weighted skip connections. For the training algorithm, virtual velocity constraints and network scheduling are proposed. For the training tasks, different feature vector selections are discussed. For the implementation, aspects of gradient-free learning using 1 GPU and the handling of perturbation noise therefore are discussed. The effects of proposed methods are illustrated in experiments encoding up to 14625 motion primitives. The capabilities of tiny neural networks with as few as 10 scalar parameters when scheduled on vehicle velocity are emphasized.

Mogens Graf Plessen

Within the context of autonomous driving a model-based reinforcement learning algorithm is proposed for the design of neural network-parameterized controllers. Classical model-based control methods, which include sampling- and lattice-based algorithms and model predictive control, suffer from the trade-off between model complexity and computational burden required for the online solution of expensive optimization or search problems at every short sampling time. To circumvent this trade-off, a 2-step procedure is motivated: first learning of a controller during offline training based on an arbitrarily complicated mathematical system model, before online fast feedforward evaluation of the trained controller. The contribution of this paper is the proposition of a simple gradient-free and model-based algorithm for deep reinforcement learning using task separation with hill climbing (TSHC). In particular, (i) simultaneous training on separate deterministic tasks with the purpose of encoding many motion primitives in a neural network, and (ii) the employment of maximally sparse rewards in combination with virtual velocity constraints (VVCs) in setpoint proximity are advocated.

Mogens Graf Plessen

This paper presents a method based on linear programming for trajectory planning of automated vehicles, combining obstacle avoidance, time scheduling for the reaching of waypoints and time-optimal traversal of tube-like road segments. System modeling is conducted entirely spatial-based. Kinematic vehicle dynamics as well as time are expressed in a road-aligned coordinate frame with path along the road centerline serving as the dependent variable. We elaborate on control rate constraints in the spatial domain. A vehicle dimension constraint heuristic is proposed to constrain vehicle dimensions inside road boundaries. It is outlined how friction constraints are accounted for. The discussion is extended to dynamic vehicle models. The benefits of the proposed method are illustrated by a comparison to a time-based method.

Mogens Graf Plessen, Pedro F. Lima, Jonas Martensson et al.

This paper presents a spatial-based trajectory planning method for automated vehicles under actuator, obstacle avoidance, and vehicle dimension constraints. Starting from a nonlinear kinematic bicycle model, vehicle dynamics are transformed to a road-aligned coordinate frame with path along the road centerline replacing time as the dependent variable. Space-varying vehicle dimension constraints are linearized around a reference path to pose convex optimization problems. Such constraints do not require to inflate obstacles by safety-margins and therefore maximize performance in very constrained environments. A sequential linear programming (SLP) algorithm is motivated. A linear program (LP) is solved at each SLP-iteration. The relation between LP formulation and maximum admissible traveling speeds within vehicle tire friction limits is discussed. The proposed method is evaluated in a roomy and in a tight maneuvering driving scenario, whereby a comparison to a semi-analytical clothoid-based path planner is given. Effectiveness is demonstrated particularly for very constrained environments, requiring to account for constraints and planning over the entire obstacle constellation space.