Systems and methods for predictions of state and uncertainty of objects for a motorized mobile system
A processing system for a motorized mobile system includes at least one sensor to measure one or more kinematic states of an object and at least one processor to use at least one state estimation filter and at least one object kinematic model to predict a first kinematic state estimate of the object and output the first kinematic state estimate for use by at least one other process of the motorized mobile system. The processor uses the predicted first kinematic state estimate and the one or more measured kinematic states to determine a second kinematic state estimate of the object and uses the second kinematic state estimate as an input to the state estimation filter to predict another kinematic state estimate of the object.
1. A processing system for a motorized mobile system comprising:
at least one sensor to measure one or more kinematic states of an object proximate to the motorized mobile system; and
at least one processor to use at least one recursive state estimation filter and at least one object kinematic model to repeatedly:
predict a first kinematic state estimate of the object based on a prior knowledge of state for the object;
output the first predicted kinematic state estimate of the object from the state estimation filter for use by at least one first process of the motorized mobile system, wherein the at least one other first process causes one or more actions to be taken by the motorized mobile system based on the first predicted kinematic state estimate of the object;
receive a measured kinematic state of the object observed by the sensor;
use the first predicted kinematic state estimate of the object and the measured kinematic state of the object observed by the sensor to determine a second kinematic state estimate of the object;
use the second kinematic state estimate of the object as an input to the state estimation filter to predict another first kinematic state estimate of the object, wherein the other first kinematic state estimate of the object is output from the state estimation filter for use by the at least one first process of the motorized mobile system, wherein the at least one first process causes one or more actions to be taken by the motorized mobile system based on the other first predicted first kinematic state estimate of the object;
predict a first uncertainty estimate of state based on a prior knowledge of uncertainty of state;
output the first predicted uncertainty estimate of state from the state estimation filter for use by at least one second process of the motorized mobile system, wherein the at least one second process causes one or more actions to be taken by the motorized mobile system based on the first predicted uncertainty estimate of state;
use the first predicted uncertainty estimate of state and the measured kinematic state of the object observed by the sensor to determine a second uncertainty estimate of state;
and
use the second uncertainty estimate of state as another input to the state estimation filter to predict another first uncertainty estimate of state, wherein the other first predicted uncertainty estimate of state is output from the state estimation filter for use by the at least one second process of the motorized mobile system, wherein the at least one second process causes one or more actions to be taken by the motorized mobile system based on the other first predicted uncertainty estimate of state.
2. The processing system of claim 1 wherein the first predicted uncertainty estimate of state comprises at least one of an uncertainty in an estimate of a location of the object, an uncertainty in an estimate of a position of the object, an uncertainty in an estimate of a velocity of the object, an uncertainty in an estimate of a distance to the object, an uncertainty in an estimate of a heading of the object, an uncertainty in an estimate of a trend of the object, an uncertainty in an estimate of an altitude of the object, and an uncertainty in an estimate of an acceleration of the object.
3. The processing system of claim 1 wherein the prior knowledge of uncertainty comprises another kinematic uncertainty estimate.
4. The processing system of claim 1 wherein the prior knowledge of uncertainty comprises one or more setup values, and the processor retrieves the prior knowledge of uncertainty from memory.
5. The processing system of claim 1 wherein the prior knowledge of uncertainty comprises a setup value based on averaging a plurality of sensor measurements to a target over time, calculating an average of the sensor measurements, and determining a deviation of the sensor measurements from the average for a standard distribution.
6. The processing system of claim 1 wherein the processor determines the second uncertainty estimate of state based, at least in part, on a difference between the measured kinematic state of the object observed by the sensor and the first predicted kinematic state estimate of the object.
7. The processing system of claim 1 wherein the processor determines the second uncertainty estimate of state based, at least in part, on the first predicted uncertainty estimate of state and a Kalman gain for the motorized mobile system.
8. The processing system of claim 1 wherein the processor defines a validation region for the object, wherein the first predicted kinematic state estimate of the object is a center of the validation region and a size of the validation region is equal to the first predicted uncertainty estimate of state extending from the center of the validation region.
9. The processing system of claim 8 wherein the processor:
associates data for the object to an identification assigned by the processor for the object if the measured kinematic state of the object is within the validation region for the object; and
either associates the data for the object to a new identification assigned by the processor for the object or rejects the data for the object if the measured kinematic state of the object is outside of the validation region for the object.
10. The processing system of claim 9 wherein the data comprises data from the sensor.
11. The processing system of claim 9 wherein the data comprises measurement data from the sensor.
12. The processing system of claim 9 wherein the processor tracks at least one kinematic state for a plurality of objects using the state estimation filter and the at least one object kinematic model.
13. The processing system of claim 1 wherein the one or more kinematic states comprises at least one of a position, a distance, a heading, a trend, an altitude, a velocity, and an acceleration, each relative to a frame of reference of the motorized mobile system.
14. The processing system of claim 1 wherein the measured kinematic state of the object comprises at least one of a position measurement, a distance measurement, a heading measurement, a trend measurement, an altitude measurement, a velocity measurement, and an acceleration measurement.
15. The processing system of claim 1 wherein the at least one first process uses the first predicted kinematic state estimate of the object for at least one of collision avoidance, stability management, a navigation operation, and an alert operation.
16. The processing system of claim 1 wherein the first predicted kinematic state estimate of the object comprises at least one of an estimate of a location of the object, an estimate of a position of the object, an estimate of a velocity of the object, an estimate of a distance to the object, an estimate of a heading of the object, an estimate of a trend of the object, an estimate of an altitude of the object, and an estimate of an acceleration of the object.
17. The processing system of claim 1 wherein the prior knowledge of state comprises another kinematic state estimate.
18. The processing system of claim 1 wherein the prior knowledge of state comprises one or more setup values, and the processor retrieves the prior knowledge of state from memory.
19. The processing system of claim 1 wherein the processor averages the first predicted kinematic state estimate of the object and the measured kinematic state of the object observed by the sensor to result in the second kinematic state estimate of the object.
20. The processing system of claim 1 wherein the processor determines the second kinematic state estimate of the object based, at least in part, on the first predicted kinematic state estimate of the object, the measured kinematic state of the object observed by the sensor, and a Kalman gain for the motorized mobile system.
21. The processing system of claim 1 wherein the processor determines the second kinematic state estimate of the object by adding the first predicted kinematic state estimate of the object to a Kalman gain for the motorized mobile system multiplied by a quantity of the first predicted kinematic state estimate of the object times an output matrix subtracted from the measured kinematic state of the object observed by the sensor.
22. The processing system of claim 1 wherein the processor selects the at least one object kinematic model from a plurality of models to match a movement of the object.
23. The processing system of claim 1 wherein the processor executes the state estimation filter that uses the object kinematic model, and the state estimation filter comprises at least one of a recursive filter, a Kalman filter, and a joint probabilistic data association filter.
24. The processing system of claim 1 wherein the object kinematic model comprises one or more algorithms.
25. The processing system of claim 1 wherein the object kinematic model comprises at least one of a maneuver model of a target, a constant velocity model, a down range acceleration model, and a cross range acceleration model.
26. The processing system of claim 1 wherein the object kinematic model comprises at least one of an interacting multiple model and a dynamically structured interacting multiple model.
27. The processing system of claim 1 wherein the object comprises at least one of a thing, a person, an animal, a ground feature, and a surface condition.
28. The processing system of claim 1 wherein the sensor comprises at least one of an optical sensor, a sound sensor, a hall effect sensor, a proximity sensor, a radar sensor, a sonar sensor, an ultrasonic sensor, a LIDAR sensor, a stereo vision camera, a structured light camera, and a time of flight camera.
29. The processing system of claim 1 wherein the motorized mobile system consists of one or more of a mobility scooter, an electronic conveyance vehicle, a riding lawn mower, a grocery cart, an all-terrain vehicle, an off-road vehicle, and a golf cart.
30. The processing system of claim 1 wherein the motorized mobile system is a mobile chair.