【摘要】：為建立穩(wěn)定可靠的分解爐溫度預測模型,結(jié)合與分解爐溫度密切相關的幾個主要運行參數(shù),提出一種粒子群參數(shù)優(yōu)化的支持向量回歸機算法(PSO-SVR),并在粒子群算法中引入自適應慣性權重的思想,構(gòu)建出分解爐溫度預測模型。與未改進的模型進行仿真對比實驗,實驗結(jié)果表明,該IPSO-SVR模型具有較佳的預測能力,預測相關系數(shù)達到0.707 5,溫度預測誤差絕對值不超過7℃,誤差率在0.8%以內(nèi)。
[Abstract]:In order to establish a stable and reliable calciner temperature prediction model, combined with several main operating parameters closely related to calciner temperature, A support vector regression algorithm (PSO-SVR) for particle swarm optimization (PSO) is proposed, and the adaptive inertial weight is introduced into PSO to construct the temperature prediction model of calciner. Compared with the unimproved model, the experimental results show that the IPSO-SVR model has better prediction ability, the prediction correlation coefficient is 0.707, the absolute value of temperature prediction error is less than 7 鈩，

本文編號：2282711