Heuristic method of automated and learning control, and building automation systems thereof

• US 10,705,492 B2
• Filed: 06/12/2018
• Issued: 07/07/2020
• Est. Priority Date: 06/12/2017
• Status: Active Grant

• Current Assignee: PassiveLogic, Inc.
• Original Assignee: PassiveLogic, Inc.
• Inventors: Harvey, Troy Aaron
• Primary Examiner(s): Patel, Ramesh B
• Application Number:
  US16/006,715
• Publication Number:
  US 20180356775A1
• Time in Patent Office: 756 Days
• Field of Search: None
• CPC Class Codes: F24F 11/63 Electronic processing
• CPC Class Codes: G05B 13/0265 the criterion being a learning criterion
• CPC Class Codes: G05B 13/048 using a predictor
• CPC Class Codes: G05B 15/02 electric
• CPC Class Codes: G05B 17/02 electric
• CPC Class Codes: G05B 2219/2642 Domotique, domestic, home control, automation, smart house

Abstract

Apparatuses, systems, and methods of physical-model based building automation using in-situ regression to optimize control systems are presented. A simulation engine is configured to simulate a behavior or a controlled system using a physical model for the controlled system. A data stream comprises data from a controlled system. A training loop is configured to compare an output of a simulation engine to a data stream using a heuristic so that a physical model is regressed in a manner that the output of the simulation engine approaches the data stream.

Claims

1. A building system controller comprising:

a controlled system;

a physical model for the controlled system, the physical model comprising a mathematical model of the controlled system;

a simulation engine configured to simulate a behavior of the controlled system using the physical model for the controlled system;

a heuristic;

a data stream, comprising data from the controlled system; and

a training loop configured to compare an output of the simulation engine to the data stream using the heuristic, such that the physical model is regressed in a manner that the output of the simulation engine approaches the data stream.

2. The building system controller of claim 1, wherein the regressed physical model is utilized by the simulation engine to more accurately predict a future behavior of a control loop for the controlled system, such that an optimal control regime is accurately computed.

3. The building system controller of claim 2, wherein the optimal control regime minimizes a cost function such that the control loop controls the controlled system with a least expense, according to the cost function.

4. The building system controller of claim 2, wherein the optimal control regime is selected through comparison of one or more potential control regimes, wherein the comparison is performed by one or more of the following but not limited to:

differential comparison, multivariate population selection, statistical classification, clustering, feature extraction, preference ranking, and benchmarking.

5. The building system controller of claim 1, wherein the regressed physical model is utilized to detect one or more faults, by detecting divergent parameters through regression of a previously regressed model.

6. The building system controller of claim 1, further comprising a control loop that controls the controlled system.

7. The building system controller of claim 1, further comprising a control loop that controls the controlled system using the output of the simulation engine to predict a future behavior of the controlled system under an arbitrary control regime.

8. The building system controller of claim 7, wherein the control loop applies the arbitrary control regime to the controlled system and observes a response of the controlled system via the data stream, thereby completing the control loop.

9. The building system controller of claim 7, wherein the control loop utilizes a cost function of the physical model to evaluate a cost of the arbitrary control regime.

10. The building system controller of claim 9, wherein the cost of the arbitrary control regime is evaluated in terms comprising one or more of, but not limited to:

energy use, primary energy use, energy monetary cost, human comfort, a safety of a building or building contents, a durability of a building or building contents, microorganism growth potential, system equipment durability, system equipment longevity, environmental impact, and energy use CO2 potential.

11. The building system controller of claim 1, wherein the controlled system comprises one or more of the following:

building automation systems, heating systems, cooling systems, ventilation systems, power management systems, renewable energy systems, irrigation systems, occupancy systems, lighting systems, environmental control systems, humidity control systems, air quality management systems, window operators, and shade systems.

12. The building system controller of claim 1, wherein the physical model is regressed using one or more of the following, but not limited to:

differential comparison, multivariate population selection, statistical classification, clustering, feature extraction, preference ranking, and benchmarking.

13. The building system controller of claim 1, wherein an extent to which the simulation engine output approaches the data stream comprises, but is not limited to:

a convergence of predicted and observed values;

a reduction in error between predicted and observed values to within an arbitrary threshold;

a reduction in uncertainty of predicted values to within an arbitrary threshold;

reaching an arbitrary threshold on number of erroneous predictions;

reaching an arbitrary threshold on number of accurate predictions;

reaching an arbitrary threshold on number of data points processed; and

reaching an arbitrary threshold on computational time spent processing data.

14. The building system controller of claim 1, wherein the output of the simulation engine is compared to the data stream using the heuristic by one or more of, but not limited to:

data slicing, time slicing, time windowing, time batching, parameter slicing, parameter windowing, single-point, and multi-point.

15. The building system controller of claim 1, wherein the heuristic comprises one or more of, but is not limited to:

decomposition methods, inductive methods, reduction methods, constructive methods, and local search methods.

16. The building system controller of claim 1, wherein the heuristic comprises one or more of a heuristic, a metaheuristic, and a hyperheuristic.

17. The building system controller of claim 16, wherein the heuristic comprises, but is not limited to, one or more of:

particle swarm organization, self-organizing migration algorithm, neural networks, group method of data handling, differential evolution, genetic algorithm, memetic algorithm, random forest, hill climbing algorithm, simulated annealing, monte-carlo methods, random search, fuzzy-logic, arithmetic mean, geometric mean, harmonic mean, trimean, median, mode, mid-range, quadratic mean (RMS), cubic mean, generalized mean, weighted mean, linear regression, logistic regression, polynomial regression, k-means clustering, k-nearest neighbors, decision trees, perceptron, multi-layer perceptron (neural network), kernel methods, support vector machines, ensemble methods, boosting, bagging, naï

ve Bayes, expectation maximization, Gaussian mixture models, Gaussian processes, principal component analysis, singular value decomposition, reinforcement learning, Voronoi decomposition, social theory voting techniques, social theory voting concepts, social welfare functions, social choice functions, single transferrable vote, Bucklin'"'"'s rule, social decision schemes, collective utility functions, Condorcet method, Condorcet method extensions, Copeland'"'"'s rule, maximin, Dodgson'"'"'s rule, Young'"'"'s rule, and ranked pairs.

18. The building system controller of claim 1, wherein the data stream comprises one or more of, but is not limited to:

a data store and a real-time data feed.

19. The building system controller of claim 1, wherein the data stream comprises one or more of, but is not limited to:

sensor data, actuation data, system parameters, equipment state, weather data, environmental data, occupant input, and occupant behavior.

20. The building system controller of claim 1, wherein the training loop regresses the physical model by comparing the output of the simulation engine to the data stream using the heuristic to improve the physical model, by one or more of varying model parameters and adapting an underlying structure and composition of the physical model.

21. A system comprising:

a coupled building system;

a data stream;

at least one memory; and

one or more processors operatively coupled to the data stream and the one or more processors operatively coupled to the at least one memory, wherein the one or more processors are configured to execute program code stored in the at least one memory to;

control the coupled building system;

receive sensor measurements obtained from the data stream;

store the sensor measurements in the at least one memory;

compute a mathematical model that substantially imitates a behavior of the controlled system, generating a predicted data stream measurement from the mathematical model;

analyze a variation between the stored sensor measurements and the predicted data stream measurement, using a cost function;

iterate a process of computation and analysis, by changing parameters of the mathematical model to regress the mathematical model such that a subsequently generated predicted data stream measurement from the regressed mathematical model approaches a subsequently received sensor measurement from the data stream; and

control the coupled building system using predicted actuation from the regressed mathematical model.

22. The system of claim 21, wherein the mathematical model is computed repeatedly, in arbitrary order, to evaluate each of an arbitrary number of different physical models.

23. The system of claim 21, wherein the one or more processors comprise one or more computational entities in a network of simulation engines.

24. The system of claim 21, wherein the one or more processors comprise collocated computational entities as well as remote computational entities.