Motor State Inference Under Noise Frequency Interference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing state estimation devices, such as abnormality diagnosis devices, can only detect noise frequency components that are greatest common divisors and integer multiples thereof, failing to consider the influence of other noise frequency components, leading to erroneous estimation.
Innovation Solution
A state estimation device that performs an estimation process based on the magnitude of a specific frequency component and restricts the estimation or operation of a motor when noise frequency components satisfy a predetermined relationship, using expressions to identify frequencies that may affect the specific frequency component, thereby suppressing erroneous estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the state estimation device detects only noise frequency components that are greatest common divisors and integer multiples thereof, then the device complexity is reduced and ease of operation is improved, but the measurement precision deteriorates and erroneous estimation occurs
Solution Approach 1:
The patent pre-calculates and stores multiple candidate noise frequency components before state estimation using the expressions: fn = |M×fs - N×f0|, |M×fs - N×fc ± 6N×f0|, and |M×fs - |fc ± 6N×f0||. This preliminary identification of all possible noise frequencies allows the estimation process to account for them without real-time calculation complexity, thereby improving measurement precision while maintaining manageable device complexity
Solution Approach 2:
The patent introduces mathematical expressions as intermediaries to identify noise frequency components. These expressions act as mediators between the sampling frequency, carrier frequency, and electrical angular frequency to systematically generate all possible noise frequencies. This intermediary approach transforms the complex problem of noise identification into a structured calculation process, improving estimation accuracy without proportionally increasing device complexity
2Reliability
If the device considers all possible noise frequency components using the given expressions, then the measurement precision is improved, but the device complexity and calculation load increase
Solution Approach 1:
The patent performs noise frequency identification in advance using predetermined expressions involving natural numbers M and N. By calculating all possible noise frequencies before state estimation and storing them as reference data, the system ensures reliable state estimation without imposing heavy real-time calculation loads, thus improving reliability while controlling device complexity
Solution Approach 2:
The patent dynamically adjusts the estimation process based on operating conditions. The control unit selects and applies appropriate expressions from among multiple candidates depending on the specific operating state, allowing the system to maintain high reliability across different conditions while optimizing calculation efficiency and managing device complexity
3Measurement precision
If the estimation process is restricted when noise frequency components satisfy a predetermined relationship, then erroneous estimation is suppressed, but the productivity and response time deteriorate
Solution Approach 1:
The patent applies preliminary anti-action by identifying noise frequency components in advance and establishing predetermined relationships between them and the specific frequency component. When such relationships are detected, the estimation process is restricted or corrected before erroneous results can occur. This approach maintains high estimation accuracy while minimizing the impact on productivity by only restricting estimation under specific noise conditions rather than continuously
Solution Approach 2:
The patent extracts and separately handles cases where noise frequency components satisfy predetermined relationships with the specific frequency component. By isolating these problematic cases and applying specific restriction or correction measures only to them, the system maintains high estimation accuracy for affected cases without unnecessarily slowing down the estimation process for normal cases, thus balancing measurement precision and productivity
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
In an estimation process, a control unit (31) estimates a state of a device (70), based on a specific frequency component (C1) of a physical quantity obtained from the device (70). In a restriction process, the control unit (31) restricts the estimation process or the operation of the motor (50) when a noise frequency component (Cn) included in the physical quantity and the specific frequency component (C1) satisfy a predetermined relationship. When a carrier frequency is represented by fc, a frequency for sampling the physical quantity is represented by fs, and an electrical angular frequency is represented by fa, a frequency of a noise frequency component (Cn) is given by any one of Expressions 1 to 3 when the physical quantity is a DC signal, and is given by any one of Expressions 4 to 6 when the physical quantity is an AC signal: fc±6Nf01Mfs−6Nf02Mfs−fc±6Nf03f0±fc±6Nf04f0±Mfs−6Nf05f0±Mfs−fc±6Nf06 where M and N are natural numbers.