Research Article
Sensors Anomaly Detection of Industrial Internet of Things Based on Isolated Forest Algorithm and Data Compression
Algorithm 1
Sensor data compression algorithm.
| Input: data.txt sensor data T number of packets processed in a single group K, error threshold E. | Output: out1.txt, out2.txt. | (1) | for i = 1 to N | (2) | read the data from “test.txt”, and write them to “data.txt” | (3) | if e of the “test.txt” | (4) | break | (5) | end if | (6) | for i = 1 to N | (7) | read the data from “data.txt” to T[i + 1] | (8) | aver = sum(T)/i + 1; | (9) | end | (10) | if (aver < 0) | (11) | for i = 1 to k | (12) | aver < aver + T[i] | (13) | aver < aver/k | (14) | end | (15) | else | (16) | for i = 2 to n | (17) | temp < aver | (18) | for j = 0 to k − 1 | (19) | if i + j ≥= n | (20) | temp < −1 | (21) | aver < aver + T[i + j] | (22) | end if | (23) | end | (24) | end if | (25) | end | (26) | end if | (27) | aver < aver/k | (28) | if |aver-temp| >= e | (29) | put i + j −1 to “out1.txt” | (30) | put T[i + j −1] to “out2.txt” | (31) | end if | (32) | return “out1.txt”,“out2.txt” |
|