FIELD: gasoline octane number measuring.
SUBSTANCE: invention is intended for measuring the octane number of gasoline. The substance of the invention lies in the fact that the effect of ultrasound on gasoline is carried out and the result of this effect is compared with the result of the same effect on gasoline with a predetermined octane number, while gasoline is exposed to ultrasound with a fixed frequency in the range from 20 to 44 kHz with the formation in it of cavitation bubbles, fix, transform and digitize at least one thousand two hundred images of cavitation bubbles of gasoline with a predetermined octane number, and based on the data obtained, machine learning is carried out, into the trained machine learning model containing the characteristic features of images of cavitation bubbles of gasoline with a predetermined octane number, at least one hundred images of cavitation bubbles formed in the tested gasoline with an unknown octane number, obtained when exposed to ultrasound of the same frequency and processed in an identical way, as a result of comparing the features of each of the entered images with the characteristic acting signs of images inherent in gasolines with a known octane number, the value of the octane number of the checked gasoline is determined by the maximum of this number of images, the number of matches of these signs.
EFFECT: ensuring the possibility of measuring the octane number of gasoline, excluding the contact of gasoline with a source of pollution, which affects the accuracy of measurements.
5 cl
| Title | Year | Author | Number |
|---|---|---|---|
| METHOD OF DETERMINING CONCENTRATION OF CORROSION INHIBITORS IN FORMATION WATER (VERSIONS) | 2024 |
|
RU2823917C1 |
| SYSTEMS AND METHODS OF ASSESSING THE VIABILITY OF EMBRYOS | 2018 |
|
RU2800079C2 |
| FUEL COMBUSTION MODES MONITORING SYSTEM BY MEANS OF TORCH IMAGES ANALYSIS USING CLASSIFIER BASED ON CONVOLUTIONAL NEURAL NETWORK | 2018 |
|
RU2713850C1 |
| METHOD OF RAISING OCTANE NUMBER OF STRAIGHT-RUN GASOLINE | 2000 |
|
RU2186825C2 |
| DEVICE AND METHOD OF THE OCTANE NUMBER DETERMINATION AND IMPURITY OF IMPURITIES IN GASOLINE WITH ULTRASOUND DIAGNOSTICS | 2017 |
|
RU2654836C1 |
| SPLICING SITES CLASSIFICATION BASED ON DEEP LEARNING | 2018 |
|
RU2780442C2 |
| ESTIMATION OF THE THICKNESS OF THE CARDIAC WALL BASED ON ECG RESULTS | 2020 |
|
RU2767883C1 |
| DEVICE FOR DETERMINING OCTANE NUMBER OF MOTOR PETROL | 2003 |
|
RU2240548C1 |
| METHODS FOR TRAINING DEEP CONVOLUTIONAL NEURAL NETWORKS BASED ON DEEP LEARNING | 2018 |
|
RU2767337C2 |
| GUIDED HEART ABLATION USING MACHINE LEARNING (ML) | 2020 |
|
RU2779871C2 |
