Spectral Power Distribution Classification Using Fuzzy Logic
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current horticultural lighting systems lack a standardized method to classify spectral power distributions (SPDs), leading to uncertainty about the effectiveness of different light sources for specific plant species, as generic terms like 'blue' and 'red' light are used without clear definitions, and existing proposals suffer from waveband overlap and broad definitions.
Innovation Solution
A method and system using radial basis functions to measure and classify SPDs, involving spectral sensors, fuzzy if-then rules, and processors to determine output classes, allowing for precise classification of light sources based on their spectral characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If generic color terms such as 'blue' and 'red' are used to characterize spectral power distributions, then the classification is simple and easy to understand, but the precision and reliability of spectral characterization is insufficient
Solution Approach 1:
The patent segments the continuous spectral power distribution into discrete wavelength bands (e.g., 300-399 nm, 400-499 nm, etc.), assigning each band a specific numerical value. This segmentation transforms the complex continuous spectrum into a structured set of discrete parameters that are both precise and systematically organized, resolving the contradiction between simplicity and precision.
Solution Approach 2:
The patent changes the parameter representation from vague color terms to precise numerical values for each wavelength band. By defining specific parameter ranges and their corresponding biological effects, the system maintains ease of use through standardized parameters while achieving high precision in spectral characterization.
2Measurement precision
If detailed spectral power distribution data is provided for each light source, then the spectral characterization is precise and reliable, but the complexity of specification and comparison increases significantly
Solution Approach 1:
The patent divides the full spectral range into standardized wavelength bands with defined boundaries. Each band is assigned a specific numerical parameter, transforming complex continuous spectral data into a manageable set of discrete values. This segmentation reduces specification complexity while preserving essential spectral information for biological effect assessment.
Solution Approach 2:
The patent transforms detailed spectral power distribution curves into a simplified parameter set representing key wavelength bands. This parameter transformation maintains precision by capturing the most biologically relevant spectral features while dramatically reducing the complexity of specification, measurement, and comparison between different light sources.
3Ease of manufacture
If waveband definitions are broad and overlapping (as in existing proposals), then the classification system is simple to implement, but the reliability and accuracy of spectral classification deteriorates
Solution Approach 1:
The patent implements non-overlapping wavelength band segments with clearly defined boundaries (e.g., 300-399 nm, 400-499 nm). Each segment is assigned a unique numerical identifier, eliminating the ambiguity caused by overlapping wavebands. This segmentation approach maintains implementation simplicity while significantly improving classification reliability through unambiguous spectral region assignment.
Data Source
AI summary
A means to automate, using fuzzy logic, the classification of spectral power distributions of optical radiation for lighting systems, and more particularly horticultural lighting systems, is presented. After inputting the spectral power distribution of optical radiation from one or more light sources, radial basis function weights for the spectral power distribution are determined and fuzzified preparatory to fuzzy logic classification. Fuzzy if-then rules are then applied, and an aggregate of the rule votes from the fuzzy if-then rules applied is used to classify the spectral power distribution. The system utilizes a spectral sensor, a fuzzifier module, a fuzzy rule database, fuzzy rule engine, an output fuzzifier module, and a means of displaying the spectral power distribution classification.