Method for searching for a minimum of a multi-dimensional surface
AssigneeDana Automotive Systems Group, LLC
InventorsJohannes van Ginkel, Ben Goyens
Life of this patent
Estimating the term…
Abstract
Provided herein is a method for calibrating a clutch by searching for the minimum of a multi-dimensional surface including determining the error between a spline function and recorded data relating to clutch characteristics, creating a multi-dimensional surface corresponding to the error values, determining the minimum of the multi-dimensional surface using the steps of performing a Steepest Gradient & Direction determination step and conducting a Golden Section Search and Switch Direction Step to find a minimum that meets a predetermined closing condition. Additionally, provided herein is a computer-implemented system for calibrating the clutch.
Claims12
A method for calibrating a clutch by searching for the minimum of a multi-dimensional surface, comprising:
- executing a plurality of steps by a computer-implemented system, wherein the steps comprise:
- obtaining a set of measured characteristics of the clutch over an actuation sequence;
- determining a set spline functions to fit the measured characteristics;
- determining error values between the spline functions and the measured characteristics;
- creating a multi-dimensional surface representing the error values;
- calculating a gradient of the multi-dimensional surface in zx, zy, and z directions in a boundary area;
- determining a direction of a steepest gradient of the gradients of the multi-dimensional surface in the zx, zy, and z directions;
- defining an initial line using the steepest gradient as a slope, a starting point and an end point;
- performing a first Golden Section Search step to identify an initial minimum;
- defining a second line using the initial minimum and switching the direction of the slope of the initial line;
- performing a second Golden Section Search step to identify a second minimum;
- calculating a distance between the initial minimum and the second minimum;
- comparing the distance to a predetermined condition;
- proceeding in an iterative manner in defining a new line using the second minimum and switching a direction of slope of the second line and performing a third Golden Section Search until the predetermined condition is met;
- returning an overall minimum;
- identifying a best fit spline function that corresponds to the overall minimum;
- and
- using the best fit spline function to calibrate the clutch.
A computer-implemented system for calibrating a clutch, the computer-implemented system comprising:
- a digital processing device comprising an operating system configured to perform executable instructions and a memory device;
- a computer program including instructions executable by the digital processing device, the computer program comprising a software module configured to manage an operation of the clutch;
- a control system including a control unit and an actuator in communication with the clutch;
- and
- a plurality of sensors in communication with the control system configured to monitor clutch characteristics;
- and
- wherein the software module includes a plurality of software sub-modules configured to determine optimal operational conditions for the clutch based at least in part on one of the clutch characteristics monitored by the plurality of sensors, and wherein the software module further comprises a module configured to fit the monitored clutch characteristics values to spline functions and a module configured to produce a multi-dimensional surface representing error values between the spline functions and the monitored clutch characteristics.
Dependencies are inferred from the claim text.