Resource Allocation Using Differential Evolution Under Nonlinear Constraints

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing resource allocation systems struggle with accurately analyzing non-linear constraints and providing flexible, easy-to-understand results, especially in complex scenarios like insurance portfolios, where traditional quadratic algorithms fail to account for asymmetrical risks and under-diversification.

Innovation Solution

A differential evolutionary algorithm is employed to optimize resource allocations, combined with resampling processes using non-parameterized historical data and moment matching, to generate robust and diversified portfolio solutions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional quadratic algorithms are used for resource allocation, then the system can handle linear constraints efficiently, but it fails to accurately analyze non-linear constraints and produces under-diversified results

Engineering Contradiction:

Improveaccuracy of non-linear constraint analysisVSAvoidability to handle diverse constraint types

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent transforms the optimization problem by changing the parameter representation from direct resource allocation to probability distributions. The differential evolution algorithm operates on distribution parameters (mean, variance, skewness, kurtosis) rather than direct allocation values, enabling accurate handling of non-linear constraints while maintaining computational efficiency. This parameter transformation allows the system to capture asymmetrical risks and produce diversified results.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces traditional mechanical optimization algorithms (quadratic programming) with a biological inspiration-based differential evolution algorithm. This substitution enables the system to handle non-linear constraints effectively by using population-based search and mutation operations that can navigate complex constraint landscapes, overcoming the limitations of deterministic quadratic methods.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If complex optimization algorithms are used to improve accuracy, then non-linear constraints can be handled, but the results become difficult to interpret and understand

Engineering Contradiction:

Improveaccuracy of resource allocation analysisVSAvoidease of understanding results

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The patent creates simplified copies or representations of the complex optimization results through moment matching. By matching the first four moments (mean, variance, skewness, kurtosis) of the optimized allocation distributions, the system generates tractable approximate solutions that retain the essential characteristics of the complex results while being easier to interpret and communicate to stakeholders.

Inventive Principle:

Principle #26Copying

3Productivity

If traditional optimization methods are used, then computational speed may be maintained, but the results suffer from model sensitivity and under-diversification

Engineering Contradiction:

Improvecomputational speedVSAvoidrobustness against model sensitivity

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent introduces dynamics into the optimization process by using probability distributions rather than fixed allocation values. The differential evolution algorithm iteratively evolves distribution parameters, allowing the system to adapt to non-linear constraints dynamically. This dynamic approach produces more robust results that are less sensitive to model assumptions while maintaining computational efficiency through the structured evolution process.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent combines multiple optimization techniques and representations into a composite approach. It integrates differential evolution algorithm with moment matching, combines distribution-based representation with traditional optimization goals, and merges handling of multiple constraint types. This composite methodology achieves both computational speed and result robustness by leveraging the strengths of each component.

Inventive Principle:

Principle #40Composite materials

Data Source

PatentUS12561744B2Differential evolution algorithm to allocate resources

Publication Date: 2026.02.24 HARTFORD FIRE INSURANCE CO

AI summary

Some embodiments are directed to a resource allocation analysis system implemented via a back-end application computer server. A resource data store may contain electronic records associated with a set of resource types, each electronic record including an electronic record identifier and resource parameter. The back-end application computer server may receive, from the resource data store, information about a set of resource types to be analyzed, including the associated resource parameters. The computer server may then execute a differential evolutionary algorithm to optimize the set of resource types based on at least one non-linear constraint and generate resource analysis results. The back-end application computer server may, according to some embodiments, perform a resampling process that uses non-parameterized historical data, regression on at least one resource type, and moment matching.