煤矿开采过程中常常会产生易燃易爆和有毒有害气体，例如瓦斯、一氧化碳、二氧化氮等，这些气体存在着安全隐患并严重危害着井下工作人员的生命安全。矿井通风系统通过向煤矿井下各个用风区域输送持续的、充足的新鲜空气，同时排放出有毒有害废气、粉尘，从而保证各个巷道的风量达到安全需求，避免安全事故的发生。因此，矿井风量调节是矿井通风系统的关键，对于提高生产效率以及保证井下工作人员的生命健康至关重要。

现有的矿井通风系统质量评价方法主要有层次划分法、模糊综合评价法、未确知测度理论，这些方法都比较依赖专家的主观经验；由于矿井通风网络风量调节是一个非定解、非线性问题，直接采用传统的解析法和数值迭代法易陷入局部最优且计算效率低，针对以上问题，本文从通风质量评价方法和通风网络风量调节算法两方面展开研究，主要内容如下：

合理的矿井通风系统质量评价方法是风量调节的前提和基础。现有的通风评价指标复杂，评价方法依赖专家的主观经验，针对上述问题，本文对经典的贝叶斯判别法进行了改进，提出了一种通风质量评价模型。首先根据《煤矿安全规程》等准则制定了各个用风区域的评价指标体系，并利用主成分分析法降低评价指标的冗余，避免信息叠加；然后引入变异系数法赋予客观权值，从而去量纲化并消除主观因素的影响，在贝叶斯判别法中利用变异性权值计算出多指标概率，以最大概率归属原则得出评价结果。实验结果表明，将本算法应用于五种用风需求不同的区域，其评价准确率均高于对比算法。

针对采用传统的解析法和数值迭代法易陷入局部最优且计算效率低的问题，本文提出了一种优化的矿井通风网络风量调节方法。首先，以通风网络待调节风量可调最大值作为优化目标，并将矿井通风的三大平衡定律作为约束条件，构建了矿井通风网络风量优化调节的数学模型；然后，采用麻雀优化算法（Sparrow Search Algorithm，SSA）对该目标函数进行求解。传统的麻雀优化算法存在迭代后期会出现种群多样性减少、易陷入局部最优等问题，基于混沌萤火虫扰动的麻雀算法（Chaotic Firefly Affect Sparrow Search Algorithm，FA_SSA）使用Tent混沌映射初始化种群以及利用萤火虫扰动方式对麻雀位置更新，从而避免了算法陷入局部最优，提高算法的收敛速度。实验结果表明，与其他优化算法相比，FA_SSA算法优化后风量可调节的范围更大，为风量调节提供一种思路。

Coal mining often generates flammable, explosive and toxic gases such as gas, carbon monoxide and nitrogen dioxide, which pose safety hazards and serious risks to  underground workers. The mine ventilation system delivers a continuous supply of fresh air to the underground areas of the mine, while discharging toxic and harmful exhaust gases and dust, thus ensuring that the air volume in each tunnel meets the safety requirements and preventing safety accidents from occurring. Therefore, air volume adjustment is the key to mine ventilation system, which is crucial to improve productivity and ensure the life and health of underground workers.

The existing methods of mine ventilation system quality evaluation mainly include analytic hierarchy process, fuzzy comprehensive evaluation method, and unascertained measurement theory, all of which rely on the subjective experience of experts;since mine ventilation network air volume adjustment is a non-deterministic and non-linear problem, the traditional analytical method and numerical iterative method are easy to fall into local optimum and low computational efficiency. The main contents are as follows:

(1) A reasonable method for evaluating the quality of mine ventilation system is the premise and basis of air volume adjustment. The existing ventilation evaluation indexes are complex and the evaluation method relies on the subjective experience of experts.In view of the above problems, this paper improves the classical Bayesian discriminant method and proposes a ventilation quality evaluation model. Firstly, the evaluation index system of each wind-using area is developed according to the guidelines of the Coal Mine Safety Regulations, etc., and the principal component analysis is used to reduce the redundancy of evaluation indexes and avoid information superposition;then the coefficient of variation method is introduced to assign objective weights so as to de-quantize and eliminate the influence of subjective factors, and the multi-index probabilities are calculated in the Bayesian discriminant method using the variability weights,and the evaluation results are derived by the principle of maximum probability attribution. The experimental results show that the evaluation accuracy of this algorithm applied to five regions with different wind demands is higher than that of the comparison algorithm.

(2) To address the problem that the traditional analytical and numerical iterative methods tend to fall into local optimality and low computational efficiency, this paper proposes an optimized method for air volume adjustment in mine ventilation networks. Firstly, based on the concept of graph theory, the mathematical model for optimal regulation of mine ventilation network air volume is constructed by taking the adjustable maximum value of the ventilation network to be regulated as the optimization objective and the three balance laws of mine ventilation as the constraints; then, the Sparrow Search Algorithm (SSA) is used to solve the objective function. The traditional sparrow optimization algorithm suffers from the problems of population diversity reduction and falling into local optimum at the late iteration. The Chaotic Firefly Affect Sparrow Search Algorithm (FA_SSA) uses Tent chaotic mapping to initialize the population and firefly perturbation to update the sparrow position, thus avoiding the algorithm to fall into local optimum and improving the convergence speed of the algorithm.The experimental results show that, compared with other optimization algorithms, the FA_SSA algorithm optimizes a larger range of adjustable air volume, which provides an idea for air volume regulation.

[1] 张建强,宁树正,陈美英,龚汉宏,张莉.我国煤炭资源开发前景及对策[J].地质论评,2020,66(S1):143-145.

[2] Improta G , Perrone A ,Russo M A , et al. Health technology assessment (HTA) of optoelectronic biosensors for oncology by analytic hierarchy process (AHP) and Likert scale[J]. BMC Medical Research Methodology, 2019, 19. (1):1-14.

[3] Huu-Tho N, Zawiah M, Yusoff N , et al. An Integrated MCDM Model for Conveyor Equipment Evaluation and Selection in an FMC Based on a Fuzzy AHP and Fuzzy ARAS in the Presence of Vagueness[J]. Plos One, 2016, 11(4): e0153222.

[4] 陈卓,丁利,曹天红,程云辉,文李,许宙,陈茂龙,焦叶,李虹辉.基于层次分析法和随机森林回归算法的谷物资源风险评估模型[J].食品与机械,2021,37(12):58-66.

[5] Yao X,Deng H,Zhang T,et al.Multistage fuzzy comprehensive evaluation of landslide hazards based on a cloud model[J].Plos One,2019, 14(11):e0224312.

[6] 苏淑娴,欧阳名三.基于粗糙集和改进胶囊网络的煤矿智能通风管理方法[J].煤炭科学技术,2021,49(7):124-132.

[7] 王子云,张嘉男,赵东霖,宋进.基于加权聚类分析法的矿井通风系统评价与优化研究[J].西部探矿工程,2019,31(12):78-81.

[8] 李春辉,陈日辉,苏恒瑜.灰色聚类评估模型在矿井通风系统中的应用[J].煤炭技术,2010,29(3):120-122.

[9] 郝继飞. 模糊Bayes判别法在判别归类系统中的应用研究[D].北京：中国地质大学,2015.

[10] 琚棋定,胡友彪,张淑莹.基于主成分分析与贝叶斯判别法的矿井突水水源识别方法研究[J].煤炭工程,2018,50(12):90-94.

[11] 崔光磊,熊伟.贝叶斯判别法在煤与瓦斯突出预测中的应用[J].煤炭工程,2013,45(03):96-98.

[12] 任冬梅,张宇洋,董新玲.应用于石油钻井安全评价的改进主成分分析-贝叶斯判别方法[J].计算机应用,2017,37(06):1820-1824.

[13] 许旭,谢贤平,郭宇航,黄兆兴,许秦.基于云模型理论的通风系统可靠性评价研究[J].化工矿物与加工,2020,49(01):7-11+19.

[14] Wang Y J.Characteristics of multiple-fan ventilation networks[J]. Geote-chnical and Geological Engineering, 1984, 2(3):229-243.

[15] 徐靖南,张民正.地下矿山通风反馈控制技术研究[J].有色金属(矿山部分),1998,1(4):42-44.

[16] 张京兆,常心坦,李学文.B/S结构的高性能通风计算系统研究[J].煤炭科学技术,2002,30(3):50-52.

[17] 张永平,吴兵,李先章.刘家梁矿通风系统优化改造研究[J].煤炭工程,2010,5(5):4-6.

[18] 庞大芳,高东旭,李绪萍.矿井通风系统优化改造研究[J].煤炭技术,2008,27(10):73-75.

[19] 卢新明,尹红.矿井通风智能化理论与技术[J].煤炭学报,2020,45(6):2236-2247.

[20] 陈开岩,周福宝,夏同强,魏连江.基于空气状态参数与风量耦合迭代的风网解算方法[J].中国矿业大学学报,2021,50(04):613-623.

[21] 司俊鸿. 矿井通风系统风流参数动态监测及风量调节优化[D].徐州：中国矿业大学,2012.

[22] 谢贤平,冯长根,王红绪.矿井通风网络模糊优化数学模型及其数值解法[J].中国矿业,1999,8(6):81-85.

[23] Xu G,Huang J,Nie B,et al. Calibration of Mine Ventilation Network Models Using the Non-Linear Optimization Algorithm[J]. Energies, 2018, 11(1): 1-9.

[24] Chen K,Si J,Zhou F,et al. Optimization of air quantity regulation in mine ventilation networks using the improved differential evolution algorithm and critical path method[J]. International Journal of Mining Science and Technology, 2015, 25(1): 79-84.

[25] Nyaaba W,Frimpong S,El-Nagdy K A.Optimisation of mine ventilation networks using the Lagrangian algorithm for equality constraints[J]. International Journal of Surface Mining Reclamation & Environment, 2015, 29(3):201-212.

[26] 王天乐. 矿井通风网络优化的风量渐进法研究[D].青岛：山东科技大学,2019.

[27] Johnson T B. Optimum open pit mine production scheduling. [J]. optimum open pit mine production scheduling,1986,16(5): 163-171.

[28] Mitchell D,Goddlad G.The use of safety related control systems in primary mine ventilation circulation[C]// International Conference on Computers & Safety. IEEE Xplore, 1989:8-10.

[29] Ralston Jonathon,Reid David,Hargrave Chad,Hainsworth David.Sensing for advancing mining automation capability:A review of underground automation technology development[J].International Journal of Mining Science and Technology,2014,24(3):305-310.

[30] 牛成强,张文静,王泉,等.果园风送喷雾风量调节研究现状与趋势[J].中国农机化学报,2020,41(12):48-54.

[31] 杨雪花,刘磊.矿井主通风机风量调节智能控制算法研究[J].煤矿机械,2020,41(1):31-33.

[32] 丰胜成,付华.矿井通风网络风量调节分支的优化选择[J].辽宁工程技术大学学报(自然科学版),2019,38(6):513-516.

[33] 许克南,王佰顺.风窗对风压及风量调节的影响分析[J].矿业安全与环保,2018,45(3):120-123.

[34] 崔传波,蒋曙光,王凯,邵昊,吴征艳.基于风量可调度的矿井风量调节[J].工矿自动化,2016,42(2):39-43.

[35] 高卫峰,罗宇婷,原杨飞.求解非线性方程组的智能优化算法综述[J].控制与决策,2021,36(4):769-778.

[36] Xie X, Yanping Y U.Optimization of mine ventilation system based on genetic algorithms[C]// International Symposium on Safety Science & Technology, 2008:1546-1550.

[37] 厍向阳,常新坦,孙艺珍.基于遗传算法的通风网络两步法风流调节优化算法[J].中南大学学报(自然科学版),2011,42(9):2729-2736.

[38] 吴新忠,张兆龙,程健维,胡建豪,任子晖.矿井通风网络的多种群自适应粒子群算法优化研究[J].煤炭工程,2019,51(2):75-81.

[39] 吴新忠,胡建豪,魏连江,钱晓喻,任子晖,张芝超.矿井通风网络的反向增强型烟花算法优化研究[J].工矿自动化,2019,45(10):17-22+67.

[40] Xue J,Shen B.A novel swarm intelligence optimization approach: sparrow sear-ch algorithm[J].Systems Science & Control Engineering An Open Access Jour-nal,2020,8(1):22-34.

[41] 马晨佩,李明辉,巩强令,杨白月.基于麻雀搜索算法优化支持向量机的滚动轴承故障诊断[J].科学技术与工程,2021,21(10):4025-4029.

[42] 王建新,李腾旭,王晔茹.基于离散型麻雀搜索算法的食品抽检路径优化[J].中国食品卫生杂志,2021,33(4):409-414.

[43] 许亮,张紫叶,陈曦,赵世伟,王鹿洋,王涛.基于改进麻雀搜索算法优化BP神经网络的气动光学成像偏移预测[J].光电子·激光,2021,32(6):653-658.

[44] 仝卫国,郭超宇,赵如意.基于改进麻雀算法优化LSSVM的再循环箱浆液密度预测模型[J].电子测量技术,2022,45(1):70-76.

[45] 张宏峰,倪受东,赵亮,张猛.基于麻雀搜索算法的摄像机标定优化方法[J].激光与光电子学进展,2021,58(22):384-390.

[46] 张俭让,杨日丽.主成分分析法在煤矿通风系统评价中的应用[J].西安科技大学学报,2011,(6):745-749.

[47] 赵旭. 基于用风区域的矿井通风质量智能评价方法研究与实现[D].西安：西安科技大学,2021.

[48] 秋兴国,刘杰,李娜,黄润青.改进贝叶斯判别法的矿井水源识别模型[J].西安科技大学学报,2022,42(02):237-244.

[49] 段志伟,杜立杰,吕海明,王家海,刘海东,富勇明.基于主成分分析与BP神经网络的TBM围岩可掘性分级实时识别方法研究[J].隧道建设(中英文),2020,40(3):379-388.

[50] 王亚,周孟然,闫鹏程,何晨阳,刘栋.PCA-BP模型在判别基于LIF技术煤矿突水水源的应用 [J].光谱学与光谱分析,2017,37(3):978-983.

[51] YANG, L., ZHAO, X., PENG, S. et al. Integration of bayesian analysis for eu-trophication prediction and assessment in a landscape lake[J]. Environ Monitori-ng and Assessment ,2014,187(1):1-19.

[52] 李韶慧,周忠发,但雨生,尹林江.基于组合赋权贝叶斯模型的平寨水库水质评价[J].水土保持通报,2020,40(02):211-217.

[53] 白礼虎.基于模糊互补判断矩阵和直觉模糊熵的决策研究[D].合肥:安徽大学,2013.

[54] 徐国冲,李威瑢.我国城市治理的评估与发展——基于变异系数法的聚类分析[J].发展研究,2019(9):45-57.