| 摘要 |
A decision dashboard system to balance competing objectives in a control system suggests control parameter settings for the control system based on specified weights for the competing objectives. For example, given a target amount of energy consumption for a campus-wide, heating, ventilation, and air conditioning (HVAC) control system, providing an optimal amount of climate control to the different campus locations may compete against the length of time climate control is provided to each of the locations. In some aspects, sets of control system settings may be used to predict outcomes based on a history of control parameter settings and measurements from the control system. In further aspects, the predicted outcomes and their associated control parameter settings may be mapped to different weights for the competing objectives and displayed in a graphical user interface. |
| 主权项 |
1. A method comprising:
providing, by a processing circuit, a graphical user interface (GUI) to an electronic display, wherein a screen of the GUI includes an input mechanism operable to assign one or more weights to competing objectives for a controlled system, wherein the competing objectives cannot be maximized or minimized at the same time due to a constrained resource, and wherein the screen of the GUI further includes an area in which suggested control parameter settings for the controlled system are displayed based on the one or more weights assigned to the competing objectives for the controlled system; receiving, at the processing circuit, data indicative of the one or more weights assigned to the competing objectives via the input mechanism of the screen of the GUI; generating, by the processing circuit, a plurality of unique sets of control parameter settings for the controlled system; retrieving, by the processing circuit, a history of control parameter settings used in the controlled system and resulting measurements from the controlled system; predicting, by the processing circuit, outcomes for each of the plurality of unique sets of control parameter settings for the controlled system based on the history of control parameter settings used in the controlled system and resulting measurements from the controlled system, wherein each predicted outcome corresponds to one or more measurements from the controlled system that are predicted to occur as a result of the controlled system using a particular one of the unique sets of control parameter settings; selecting, by the processing circuit, an optimal subset of the plurality of unique sets of control parameter settings based on the predicted outcomes and the competing objectives, wherein the optimal subset of the plurality of unique sets of control parameter settings corresponds to a Pareto frontier; selecting, by the processing circuit, a particular set of control parameter settings from among the optimal subset of the plurality of unique sets of control parameter settings based on the one or more weights assigned to the competing objectives; and providing, by the processing circuit, the particular selected set of control parameter settings to the screen of the GUI for presentation as the suggested control parameter settings. |
